^ X^jir^x m xfo dcx jo m x!i£Sx// ^ x^ui^x m

co

z

CO = CO — CO

"HSONIAN INSTITUTION NOIlfUliSNI NVIN0SH1IIMS S3 I a VB 8 11 LIBRARIES SMITHSONIAN IN
z ,v. co z co z <2 z

x

co .x^

I ^test

> "'^P^ 2 N'^L°£>/ > '-Sr* 2 ^ >

JosHiiiNS^SBiavaan li brar i es^smithsonian^ institution NouniusNi nvinoshiiwsX

co X co ^ co — c/>

oo

UJ

ce

<

QC

CQ

£ x^o^/ 5 o *

THS0NIANJ!NSTITUTI0N N0lini!JLSNl'JNVIN0SHl!INSZS3 I U MU 8 II ^Ll B R AR I E S'*" SM ITHSOnIaN-* I N
“ z n «. z r= z r;

co m .% 2 ra /^5\ 2 ^gjfefc co

i j§ 0 - lljl £

M0SHll!NSWS3iavaan LIBRARIES^SMITHSONIAN INSTITUTION^NOlinillSNI nvinoshiiws^s:

Z * CO Z CO

X
CO

o

>' s | g >n^? >■* ’ ' i |y

FHSONIAN INSTITUTION ^NOliniliSN^NVINOSHlIWS^SB I B VH 8 I^LI B RAR I ES^SMITHSONIAN^IN

co z \ ^ ._.■ 5 co

;4WkA K MM ij ££ H T. W?J/'

' " 5 5 2 5 m-*’-

Z _J) 2 ' =S 2

'Joshiiiajs S3 iava an libraries Smithsonian institution NouniusNi nvinoshiiiais s

z r- z £ ? rj ... 2

m xwx g m ^ xiom^ix m

THS0NIAN_INSTITUTI0N N0SlfU!iSm”NV!N0SHIilAJ$ S3 I HVH 8 ll~LI B RAR 1 ES°>SMITHSONIAN”lN

2 .V. CO Z « z W 2

- vx#S\v' ? . y- / * ♦*= vS^VL ~ 5

z Am-S -i Ay-* \ h .,*/ a* s =s ^swsks. z .

2 o fjC' re - o #*S> X. 2

co co oo &WA wW co 'fflv*. co

o wLgm X o WyWfi X mm, o

z w;-& t v^4l»7 Z Wim? t= 2 \iwfl

NOSHlIWS^SaiaVaan^LIBRARIES^SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSHiWS^S:
co ^ ___ co ^ co __ — ; .,„ . co

yj

tc

<

£C

CQ

O X£<dcx iz XOjg^X Q

THSOnIan”* INSTITUTION*^ NOlinillSNI "'nVINOSHIIINS^SB I HVa 0 n ~ 1 L I B RAR I ES^ SMITHSON lAN^IN
r- z r- » z r- Z

O y<^TTro>v vV^. O /v^v O

m ^ xjvas>^x rn ^ x^va^x m X^osA^X- g m

NosHiiiNs^sa i avaa n~n brar i es SMiTHS0NiAN""iNSTiTUTi0NWN0iiniiiSNi”wviN0SHiiiAis s:
co z , co z « co z co

I Mf//,.. I X . I ~ . I I 1

“ ^0 g j= WiJf ? mup

m ' — ---■ -?, xtams^- rn -«* ^ xoiast*X m

*™ ^O - (/) \ ^ //}

NOliniliSNI NVIN0SH11WS S3 I BVB 8 IT LIBRARIES SMITHSONIAN INSTITUTION NOUDlliSh

X? 2_ v- . (/> rr /n

LIBRARIES SMITHSON!AN_ INSTITUTION NOliniliSNI NVINOSHIUNS^SB I B VH 0 SI 2 LI B RAR I E:

<o ~ co

O ^LEX'' _ Q '\ ~ x^oTTT^ O

NOliniliSNI NVIN0SH1IIAIS S3 I a VH a 11 LI B RAR I ES^SMITHSONIAN^INSTITUTION 2 NOIlfUllSN

r * z 2 r* ^

m ' \y z m x^vos 'ii m

^ CO C/5 ^

LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVIN0SH1IINS S3 I a Va 9 11 ~LI B RAR I ES

■» S Z * CO z co ^

1 J Jk S S ^ <

O s klfc

| | ^ >* s

N0liniliSNI_NVIN0SHIIIAIS^S3 I H ViJ a n\t B RAR I ES^SMITHSONIAN^INSTITUTION ^NOliniliSN

- * m ~ m = 05

co

H

C

^ V?
o

” “* ^ cJ 2 J

LIBRARIES ^SMITHSONIAN INSTITUTION NOliniliSNI NVIN0SH1MS S3 lavaa 11 LIBRARIES

I I " i#®

I s | s | 'Csgf' m

NOliniliSNI NVINOSHIII/VS S3 I MVHfl 11 LI B RAR I £S^SMITHSON!AN”lNSTITUTlONC/> NOliniliSN
5 £2 <2- </> ^ ... m

i .A, S xtf^x I 2

<ss? '•s^'V O W' vl ' <s=? w

LI B RAR I ES^SMlTHSONIAN^INSTITUTlON^NOfiniJlSNI^NWINOSHltlAJS^Sa I HVa a 11^ LI B RAR I ES

^ .4 W ~ Z ^

NOliniliSN I *J\JVINOSHil INS 3 I a va a n “’Ll B RAR I ES2SMITHSOmANJ|NSTITUTION 2 NOliniliSN

L ® 5 , r~ h r= 2

rn ^ 2 z ^5o^rv O O

. . .• ^ m ^ m

.1 B RAR I ES SMITHSONIAN INSTITUTION NOliniliSNI “"NVINOSHilWS S3 I HVa a IT ”l I B R AR I ES

" w - ^ w z w.2-„

— 2 sfffifcA s V

Published by

The Palaeontological Association • London
Price £43-00

THE PALAEONTOLOGICAL ASSOCIATION

(Registered Charity)

The Association was founded in 1957 to promote research in palaeontology and its allied sciences.

COUNCIL 1993-1994

President : Dr W. D. I. Rolfe, National Museums of Scotland, Chambers Street, Edinburgh EH1 1JF
Vice-Presidents'. Dr A. W. Owen, Department of Geology and Applied Geology, The University, Glasgow G12 8QQ
Dr M. E. Collinson, Department of Geology, Royal Holloway and Bedford New College,

University of London, Egham, Surrey. TW20 OEX

Treasurer. Dr T. J. Palmer, Institute of Earth Studies, University of Wales, Aberystwyth, Dyfed SY23 3DB
Membership Treasurer : Dr M. J. Barker, Department of Geology, University of Portsmouth, Burnaby Road,

Portsmouth POl 3QL

Institutional Membership Treasurer. Dr J. E. Francis, Department of Earth Sciences, The University, Leeds LS2 9JJ
Secretary. Dr J. A. Crame, British Antarctic Survey, High Cross, Madingley Road, Cambridge CB3 OET
Newsletter Reporter : Dr D. Palmer, c/o Department of Earth Sciences, Downing Street, Cambridge CB2 3EQ
Marketing Manager (Sales): Dr L. Cherns, Department of Geology, University of Wales College of Cardiff CF1 3YE
Marketing Manager (Publicity): Dr P. B. Wignall, Department of Earth Sciences, The University, Leeds LS2 9JJ
Public Relations Officer: Dr P. A. Allison, P.R.I.S., The University, P.O. Box 227, Whiteknights, Reading RG6 2AB
Newsletter Editor : Dr R. B. Rickards, Department of Earth Sciences, University of Cambridge, Cambridge CB2 3EQ
Editor of Field Guides to Fossils: Dr E. A. Jarzembowski, Bishopstone Manor South, Bishopstone, East Sussex BN25 2UD

Editors

Dr J. E. Dalingwater, School of Biological Sciences, Williamson Building, The University, Manchester Ml 3 9PL (co-opted)
Dr P. D. Lane, Department of Geology, University of Keele, Keele, Staffordshire ST5 5BG
Dr P. Doyle, School of Earth Sciences, University of Greenwich, Brewhouse Lane, London El 9NU
Dr A. R. Milner. Department of Biology, Birkbeck College, Malet Street, London WC1E 7HX
Dr C. J. Cleal, Department of Botany, National Museum of Wales, Cardiff CF1 3NP
Dr R. M. Owens, Department of Geology, National Museum of Wales, Cardiff CF1 3NP

Other Members: Dr B. M. Cox, British Geological Survey, Keyworth, Nottingham NG12 5GG
Dr M. P. Smith, School of Earth Sciences, University of Birmingham, Birmingham B15 2TT

Overseas Representatives

Argentina: Dr M. O. Mancenido, Division Paleozoologia invertebrados, Facultad de Ciencias Naturales y Museo, Paseo del
Bosque, 1900 La Plata. Australia: Dr K. J. McNamara, Western Australian Museum, Francis Street, Perth, Western
Australia 6000. Canada: Professor S. H. Williams, Department of Earth Sciences, Memorial University, St John’s,
Newfoundland A1B 3X5. China: Dr Chang Mee-mann, Institute of Vertebrate Palaeontology and Palaeoanthropology,
Academia Sinica, P.O. Box 643, Beijing. Dr Rong Jia-yu, Nanjing Institute of Geology and Palaeontology, Chi-Ming-Ssu,
Nanjing. France: Dr J.-L. Henry, Institut de Geologie, Universite de Rennes, Campus de Beaulieu, Avenue du General
Leclerc, 35042 Rennes Cedex. Iberia: Professor F. Alvarez, Departamento de Geologia, Universidad de Oviedo, Cl. Jesus
Arias de Velasco, s/n. 33005 Oviedo, Spain. Japan: Dr I. Hayami, University Museum, University of Tokyo, Hongo 7-3-1,
Tokyo. New Zealand: Dr R. A. Cooper, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt. Scandinavia: Dr R.
Bromley, Fredskowej 4, 2840 Holte, Denmark. USA : Professor A. J. Rowell, Department of Geology, University of
Kansas, Lawrence, Kansas 66044. Professor N. M. Savage, Department of Geology, University of Oregon, Eugene, Oregon
97403. Professor M. A. Wilson, Department of Geology, College of Wooster, Wooster, Ohio 44961. Germany: Professor
F. T. Fursich, Institut fur Palaontologie, Universitat, D8700 Wurzburg, Pliecherwall 1

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1994 are:
Institutional membership . . . £85 00 (U.S. $170) Student membership . . . £1 1-50 (U.S. $20)

Ordinary membership . . . £28 00 (U.S. $50) Retired membership .... £14 00 (U.S. $25)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr J. E.
Francis, Department of Earth Sciences, The University, Leeds LS2 9JJ. Student members are persons receiving full-time
instruction at educational institutions recognized by the Council. On first applying for membership, an application form
should be obtained from the Membership Treasurer: Dr M. J. Barker, Department of Geology, University of Portsmouth,
Barnaby Road, Portsmouth POl 3QL. Subscriptions cover one calendar year and are due each January; they should be
sent to the Membership Treasurer. All members who join for 1994 will receive Palaeontology, Volume 37, Parts 1-4.
Enquiries concerning back numbers should be directed to the Marketing Manager.

Non-members may subscribe, and also obtain back issues up to 3 years old, at cover price through Basil Blackwell Ltd,
Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, UK. For older issues contact
the Marketing Manager.

US Mailing: Second class postage paid at Rahway, New Jersey. Postmaster: send address corrections to Palaeontology,
c/o Mercury Airfreight International Ltd, 2323 EF Randolph Avenue, Avenel, NJ 01001, USA (US mailing agent).

Cover: Early evidence of extinction: in situ tooth of extinct mastodon contrasted with that of living elephant drawn by Jan
Van Rymsdyk to illustrate William Hunter’s Royal Society paper of 1768 (see Earth Sciences History, 3, 17, fig. 1).
Reproduced by permission of the Keeper of the Hunterian Books and MSS (Glasgow University Library, Hunterian
fig. 92.3).

lib rari£>

BURLINGIIDS: SMALL PROPARIAN CAMBRIAN
TRILOBITES OF ENIGMATIC ORIGIN

by H. B. WHITTINGTON

Abstract. The original, and additional, specimens of Burlingia hectori are described from Canada and
Sweden, B. laevis being considered synonymous with B. hectori. Westergaard’s original material of
Schmalenseeia amphionura and S. acutangula is redescribed, two cranidia included with the latter being
segregated as S.jagoi sp. nov. No burlingiid exoskeleton exceeds a length of 13 mm, most being less than half
this size. The eye lobe was close to the glabella, the two branches of the facial suture directed outward and
forward. A large rostral plate beneath the frontal area is described, the incomplete hypostome is outlined in
one specimen; the condition may have been conterminant. The exoskeleton was non-fulcrate, the sclerites
separate but apparently flexure between them limited, lacking border and pleural furrows ; lateral and posterior
doublure unknown. The pygidium of B. hectori was narrow and short; in species of Schmalenseeia the
pygidium incorporated more segments, those of the thorax having been fewer. Burlingiids were world-wide in
distribution, and occur in outer shelf to slope facies. They may belong with those trilobites in which the
conterminant condition of the hypostome was retained throughout growth.

The relatively small size of burlingiids, and their unusual facial suture, have presented intriguing
problems since they were discovered early in this century. They are Cambrian in age, yet the suture
was of a type that Beecher (1897) thought arose in the post-Cambrian. Were the small specimens
the remains of immature individuals, or was this their maximum size? This study of new and old
material shows that the distinctiveness of burlingiids lies not only in their size and peculiar suture,
but in that they had an unusual rostral plate and a thorax which lacked a horizontal hinge-line and
the facet, i.e. was non-fulcrate (for definition and discussion of this term see Whittington 1990,
p. 28). Known specimens are all relatively small in size, and the exoskeleton thin; their distribution
was world-wide in outer shelf margin or slope facies.

In 1981, I described the disarticulated type material of Schmalenseeia amphionura , and complete
exoskeletons from Newfoundland. Since then, eighteen specimens of Burlingia hectori , the first
found since Walcott’s original lot of three, were obtained by a party from the Royal Ontario
Museum, Toronto, Canada. A generous invitation from Mr David M. Rudkin to study them, and
the loan of new and old material of Burlingia and Schmalenseeia from Sweden, has enabled this re-
assessment of the morphology of species of both genera to be made from the best-preserved and
most complete examples so far known.

MORPHOLOGY, MODE OF LIFE AND AFFINITIES

Exoskeleton. The broken edges of the specimens in limestone (PI. 3, fig. 6; PI. 4, figs 1, 4), and the
similarity between internal mould and cast from external mould (Whittington 1981, pi. 2), show the
thinness of the exoskeleton. Crumpling of both limestone and shale specimens is further evidence
of this thinness. The exoskeleton was non-fulcrate (PI. 1, fig. 3; PI. 3, fig. 4), curving downward with
increasing steepness, outward from the change in slope at the axial furrow. Impressed furrows -
axial, pleural or border - are lacking. Within the axial region, occipital and glabellar furrows were
shallowly impressed or absent, articulating and inter-ring furrows barely impressed. Only in the

IPalaeontology, Vol. 37, Part 1, 1994, pp. 1-16, 4 pls.|

© The Palaeontological Association

2

PALAEONTOLOGY, VOLUME 37

Newfoundland material of S. amphionura (PI. 3, fig. 5) do glabellar furrows appear moderately
impressed, and the articulating furrows have a deep pit which formed an apodeme on the inner
surface. In my earlier study of S. amphionura (Whittington 1981, p. 598) I saw no evidence of the
articulating half ring, and argued that the exoskeleton may have been a rigid shield, possibly having
had sutures between the segments. The broken posterior edges of the cephalon and thoracic
segments in the present specimens (PI. 3, fig. 6; PI. 4, figs 1, 4, 6) reveal the short (sag.), raised
articulating half ring, and the ridge along the anterior edge of the pleura. In Canadian specimens
of B. hectori (PI. 1, fig. 4; PI. 2, fig. 1) the forwardly-convex arc impressed into the axial ring may
result from the underlying doublure of the ring, and half ring of the segment following. The
disarticulation and telescoping of individual specimens (PI. 3, figs 3, 6; PI. 4. figs 1, 6) indicate that
the segments were separate from one another and from cephalon and thorax. In particular
Canadian specimens of B. hectori (PI. 1, fig. 1) the first thoracic segment is pushed beneath the
anterior edge of the cephalon, while other entire specimens (PI. 2, fig. 3) show no displacement of
segments. The Swedish examples (PI. 2, fig. 2; PI. 3, fig. 1) of this species show such displacement.
This evidence, from specimens preserved in different ways, leads me to withdraw my argument and
conclude that the burlingiid exoskeleton was not a rigid shield, but of separate, articulated sclerites.
No facet is present in any species, nor any articulating process or socket; the ridge along the anterior
margin of the pleura acted as an articulating flange beneath the turned-up edge of the sclerite in
front. In B. hectori the ridge died out abaxially, and this may have allowed the distal portions of
the pleurae to slip one below the other, and some limited convex-upward flexure of the exoskeleton.
In the two species of Schmalenseeia considered here (PI. 3, figs 2-^1; PI. 4, figs 1-6), the ridge extends
to the margin of the pleura, and this morphology suggests that convex-upward flexure was limited.
Some limited concave-upward flexure may have been possible in all three species. No even partly
enrolled example is known. Most unusual is the lack of any evidence of a doublure in the lateral
cheek area, thorax, or pygidium. The distal edge of the exoskeleton is ill-defined and irregular in
almost all specimens, an exception being the low marginal ridge visible on a small specimen of
S. amphionura (PI. 3, figs 2-3). Evidence is described below for the presence of a rostral plate (Text-
fig. 1) beneath the frontal area of the cranidium in species of both genera. The outline of a
hypostome is preserved in one Swedish specimen (PI. 2, fig. 2), the anterior margin not defined.
Hence whether or not the hypostome was linked by a suture to the rostral plate (the conterminant
condition of Fortey and Chatterton 1988) is uncertain. Because the rostral plate extends back to the
line of the preglabellar furrow, such attachment may seem probable, but remains to be confirmed.
The boss on the anterior of the fixed cheek in 5. amphionura (PI. 3, figs 3, 5) appears to lie above
the posterolateral corner of the rostral plate, and where the tip of the anterior wing of the
hypostome may have been. Whether this boss had any association with attachment of the
hypostome is uncertain; the boss is an unusual feature.

Size. Moberg (1903, p. 96) regarded Schmalenseeia amphionura as ‘one of our smallest trilobites’.
The length (exs., measured between the level of the median edge of the cephalon and that of the tip

EXPLANATION OF PLATE 1

Figs 1-5. Burlingia hectori Walcott, 1908. Middle Cambrian, Stephen Formation; British Columbia, Canada.
I, USNM 53418, lectotype; arrow on left (pointing backward) indicates anterior margin of first thoracic
segment, arrow on right (pointing forward) position of posterior margin of cephalon; thoracic segments 5
and 14 are numbered; dorsal view; x 15. 2-3, USNM 534 1 96 ; incomplete exoskeleton; dorsal and postero-
dorsal views; x 10. 4, ROM 48463; thoracic segments 5 and 15 are numbered; arrow points to notch in
posterior margin of pygidium; frontal area of cranidium removed to expose rostral plate (r); complete
exoskeleton; x 15. 5, USNM 534 1 9<r/ ; axial region broken away; beneath frontal area of cranidium rostral
plate (r) is exposed; posterior margin arrowed on left side; incomplete and poorly preserved exoskeleton;
x 15.

PLATE 1

WHITTINGTON, Burlingia

4

PALAEONTOLOGY, VOLUME 37

of the most posterior pleura) ranges from 2-8 to 5 0 mm. That of the older S. acutangula is about
4 mm, and specimens of Burlingia hectori from British Columbia, Canada, are all between 5 and
6 mm in length, except one of 8 mm. Among burlingiids only Westergaard’s type of B. laevis
(PI. 2, fig. 2) exceeds 10 mm in length. I have not included small size in the diagnosis of burlingiids
because of the difficulty of giving meaning to such a characterization. Specimens of holaspid
Agnostina (e.g. Robison 1984; 1988) range in length from 3-10 mm, and small holaspides of other
trilobites (e.g. Whittington 1957, p. 444) lie within or below this range. Burlingiids are small only in
relation to larger holaspides of certain other trilobites.

Geographical distribution. In relation to Cambrian geography (e.g. Scotese and McKerrow 1990, fig.
4) burlingiids are known from marginal, outer shelf sites of Laurentia, Baltica, Siberia (Lazarenko
1960; Soloviev 1969) and peripheral Gondwana (Avalonia (Rushton 1978), Tasmania (Jago 1972),
South China). The occurrences of Burlingia and Schmalenseeia in South China are in the slope
biofacies (W. T. Chang, pers. comm.) - in dark muddy limestones or fine-grained shales. Oil shales
in Siberia yielded B. obscura Soloviev, 1969. In Canada and China (Chang 1988, p. 55) Burlingia
occurs with Ory otocephalus, a genus which Fritz (1990, p. 108) noted is not found in shallow water
strata in North America. The Siberian species is accompanied by other oryctocephalids. I agreed
(1981, p. 599) with Jago (1972, p. 233) in suggesting, in part because of the supposedly rigid, thin
exoskeleton, that the mode of life of S. amphionura was probably planktonic. Neither the
morphology of burlingiids, nor their wide distribution, provides compelling evidence for this view.

Origin and relationships. Stubblefield (1936, p. 429) discussed burlingiids, norwoodiids and eodiscids
as Cambrian trilobites having a proparian suture, no complete exoskeleton being more than 12 mm
long. This size he regarded as considerably below the dimensions of an adult trilobite, and suggested
that such forms either were immature, or were mature forms that had arisen by paedomorphosis.
Fortey and Owens (1990) reviewed the operation of this latter process in trilobites, and point to one
evolutionary trend, miniaturization, as occurring throughout the Palaeozoic. The species which
portray this trend are not named, but are characterized as having a mature size of a few millimetres.
They are considered to have different origins, but to be comparable adaptations. I assume that
burlingiids are included among the families they plot (figure 5.5), and since I reviewed (1981) the
problem of origins of new groups of trilobites, one new fact has emerged. This is that the smallest
protaspid stage of the Middle Cambrian ptychopariid Spencella sp. had a proparian suture (Fortey
and Chatterton 1988, text-fig. 10. 8 a-b\ based on their plate 17, figs 7, 9-10; the names Spencellal
and Bathyuriscus ? having been transposed in the explanation of the plate, as B. D. E. Chatterton
informs me (pers.'comm.)). In the largest protaspid stage (Fortey and Chatterton 1988, text-fig. 10.
la-b\ pi. 17, figs 1 6—1 9) the suture was opisthoparian, and in both stages the hypostome was
conterminant. The further development of ptychoparioids has been reviewed by Fortey (1990,
p. 546, text-fig 9), laying stress on how during the early meraspid period the hypostome became
natant. Little is known of development in early Middle Cambrian trilobites, and an immature stage -

EXPLANATION OF PLATE 2

Figs 1-4. Burlingia hectori Walcott, 1980. 1, 3-4, Middle Cambrian, Stephen Formation; British Columbia,
Canada; 2, Eccaparadoxides oelandicus zone, early Middle Cambrian; Oland, Sweden. 1, ROM 48449;
posterior margin of cephalon arrowed (pointing backward) on right, overlies first thoracic segment; segments
5 and 14 are numbered; notched posterior margin of pygidium arrowed; x 15. 2, SGU 6246, holotype of
B. laevis Westergaard, 1936 (original of his pi. 12, fig. 9); external mould, outline of hypostome impressed on
anterior portion of glabella; x 8. 3, ROM 48450; exoskeleton; notched posterior margin of pygidium
arrowed; xl5. 4, ROM 48454; crumpled, incomplete exoskeleton; external mould of rostral plate (r)
exposed beneath frontal area of cranidium, showing groove along anterior margin; x 15.

PLATE 2

WHITTINGTON, Burlingia

6

PALAEONTOLOGY, VOLUME 37

a small meraspis, for example - resembling a burlingiid is not known. The oldest species of
burlingiids, Burlingia hectori , B. obscura Soloviev, 1969, and Schmalenseeia acutangula , appeared at
about the same time in the early Middle Cambrian. They have many characters in common, which
suggest a relationship between them, their morphology (Text-fig. 1) being quite different from that
of any contemporary group. Distinctive of S. acutangula (PI. 4, figs 1-6) is the incorporation of
many more segments into a larger pygidium, apparently retaining the notched posterior outline
which is lost in the younger S. amphionura.

Henningsmoen (1951, p. 194) tentatively placed the Burlingiidae in the ptychoparioids, a
suggestion followed by Poulsen (in Moore 1959, p. 293). Bergstrom (1973, pp. 18, 40) noted some
of the peculiar characters of the thorax, considering that enrolment was not possible, and hence
(presumably) allied burlingiids with redlichioids. Fortey’s (1990) discussion of trilobite classification
stressed the importance of early stages in ontogeny and of the attachment (or otherwise) of the
hypostome. The apparent presence of a large rostral plate in burlingiids suggests that the hypo-
stomal condition may have been conterminant. If so, burlingiids may be trilobites exhibiting the
primary conterminant condition (Fortey 1990, p. 540), that is, the conterminant condition was
presumably retained throughout growth. Among such Cambrian trilobites are Redlichiida,
Corynexochida, Leiostegioidea, Damselloidea, and Odontopleurida. Relationships within this
group have yet to be resolved, and what little is known of ontogeny does not suggest whence
burlingiids may have been derived.

SYSTEMATIC PALAEONTOLOGY
Family burlingiidae Walcott, 1908

Diagnosis. Exoskeleton suboval in outline, gently convex; axial region gently convex, axial furrow
a change in slope, not impressed; pleural region non-fulcrate, doublure apparently absent, except
for broad (tr.) rostral plate beneath relatively long (sag.) frontal area; condition of hypostome
uncertain. Eye lobe close to anterior half of glabella, both branches of facial suture directed outward
and forward. Narrow, low, anterior and anterolateral cephalic border. Pleurae directed
progressively more strongly backward, no pleural furrow, no facet, ridge along anterior edge of each
pleura fitted beneath raised posterior edge of cephalon or pleura; no axial or other articulating
processes and sockets.

Distribution. Middle and early Upper Cambrian, worldwide.

EXPLANATION OF PLATE 3

Figs 1, 6-7. Burlingia hectori Walcott, 1908. Middle Cambrian, Eccaparadoxides oelandicus zone; Sweden.
1, SGU 6245; original of Westergaard (1936, pi. 12, fig. 8); Gland; thin exoskeletal layer partly broken and
weathered on the left side; irregular, broken posterior margin of cephalon (arrowed) overlying first thoracic
segment; dorsal view; x 9. 6-7. M. J. Collins collection, Jamtland. 6, thoracic segments disarticulated and
partly telescoped; segments 5, 8, 9 and 14 numbered; dorsal view; x 15. 7, exoskeleton exposed from ventral
side; notch in posterior margin of pygidium arrowed; x 15.

Figs 2-5. Schmalenseeia amphionura Moberg, 1903. early Upper Cambrian. 2-4, SGU 134; original of
Westergaard (1922, pi. 1, fig. 19); Skogsby, Oland, Sweden; latex cast from external mould; oblique right
lateral, dorsal and postero-dorsal views; thorax partly disarticulated; segments 3, 5 and 7 numbered;
x 18. 5, SM A 104876; original of Whittington (1981. pi. 2, fig. 3); basal Elliot Cove Group, beds with
Agnostus pisiformis; Little Ridge, east shore of Chapel Arm, 2-4 km south of McLeod Point, Trinity Bay,
eastern Newfoundland, Canada; internal mould of incomplete cephalon and four thoracic segments;
anterior portion of glabella and frontal area broken away to show rostral plate (r); arrow points to boss on
fixed cheek beside anterior glabellar lobe; x22-5.

PLATE 3

WHITTINGTON, Burlingia , Schmalenseeia

PALAEONTOLOGY, VOLUME 37

Genus burlingia Walcott, 1908

Type species. By monotypy; B. hectori Walcott, 1908, from the early Middle Cambrian, Stephen Formation,
of Mt Stephen, British Columbia, Canada.

Diagnosis. Glabella tapering slightly forward, rounded anteriorly, median occipital node, occipital
and glabellar furrows either absent, or in rare specimens SO, SI and S2 present as shallow lateral
depressions. Eye lobe of length (sag.) more than one-third that of glabella. Thorax of 14 (rarely 15)
segments in the type species, last pair of pleurae directed exsagittally, flanking the narrow (tr.),
parallel-sided pygidium, which has the short axis poorly defined, two pairs of backward-directed
pleurae behind it, posterior margin deeply notched; length (exs.) of pygidium less than that of
posterior thoracic pleurae.

Distribution. British Columbia, Canada; Sweden; northern Siberia; Guizhou province, China.
Stratigraphical range. Early Middle Cambrian.

Burlingia hectori Walcott, 1908
Plates 1-2; Plate 3, figures 1, 6-7; Text-figure 1

Lectotype (selected by Rasetti 1951, p. 138). USNM 53418. flattened exoskeleton, original of Walcott 1908,
pi. 1, fig. 8.

Paratypes. USNM 53419a-6, two specimens, figured herein, from Walcott’s collection.

Other material. ROM 48449-48467, complete and incomplete exoskeletons, crumpled and flattened.

Locality and horizon. Locality S8D of Rasetti (1951, p. 128), equivalent to USNM locality 14S, southwest flank
of Mt Stephen, 1-5 miles east 30° south of Field station. ROM locality EST (NTS 82N8, UTM 377933)
approximately 350 m. NE of locality S8D, at about 1981 m altitude, Mt Stephen, collected in 1982 and 1984.
Early Middle Cambrian, Ogygopsis klotzi faunule (Rasetti 1951, p. 101) at base of Bathyuriscus - Elrathina
zone, Stephen Formation (Fritz 1971, fig. 6; 1990, fig. 1). All the specimens are on the brown, weathered
surfaces of pieces of the dark grey shale ; none has been split from unweathered shale, so no counterparts are
available. On the surface of occasional pieces (USNM 534196, ROM 48454, 48458) are entire exoskeletons, or
scattered parts of varying size, of Oryctocephalus reynoldsi Reed, 1899. Reed illustrated his new species by a
drawing; additional specimens have been figured by Rasetti (1951) and Shergold (1969).

Description. Exoskeleton oval in outline, axial region of low convexity, shallow axial furrow, maximum width
(tr.) at about midlength of exoskeleton, and about 0-7 of length to tip 13th thoracic pleura. Genal and pleural

EXPLANATION OF PLATE 4

Figs 1-6. Schmalenseeia acutangula Westergaard, 1948. early Middle Cambrian; Scania, Sweden. 1, SGU 6354,
holotype; incomplete exoskeleton; original of Westergaard (1948, pi. 1, fig. 2); arrow indicates division
between thorax and pygidium; dorsal view; x 15. 2-3, 6, SGU 6353; incomplete exoskeleton; original of
Westergaard (1948, pi. 1, fig. 3); left lateral, anterior and dorsal views; x 15. 4-5, SGU 6355; thorax and
pygidium; original of Westergaard (1948, pi. 1, fig. 4); arrow indicates division between them; dorsal and
postero-dorsal views; x 15.

Figs 7-10, Schmalenseeia jagoi sp. nov. early Middle Cambrian; Scania, Sweden. 7, 9, SGU 6357, holotype;
original of Westergaard (1948, pi. 1, fig. 6); incomplete exfoliated cranidium; arrow indicates broken
posterior margin of cheek; dorsal and oblique right lateral views; x 12. 8, 10, SGU 6356; original of
Westergaard (1948, pi. 1, fig. 5); exfoliated cranidium; antero-dorsal and dorsal views; x 15.

PLATE 4

WHITTINGTON, Schmalenseeia

10

PALAEONTOLOGY, VOLUME 37

text-fig. 1 . Burlingia hectori Walcott,
1908. Restoration of exoskeleton, based
on Canadian and Swedish specimens, a,
ventral view of cephalon, rostral plate
and hypostome (incomplete anteriorly)
in heavier outline; doublure not known
elsewhere, except that of occipital ring.
b-c, right lateral and dorsal views. Scale
bar represents 1 mm, for largest
Canadian specimen known, for the
Swedish material it is three-fifths of this
length.

regions (PI. 1, figs 2-3) curved downward from axial furrow, no inner horizontal portion or fulcrum. Glabella
widest at base, tapers gently forward and rounded anteriorly, length (sag.) about two-thirds that of cephalon;
low median occipital node visible in most specimens; faint lateral depressions (PI. 1, fig. 4) of occipital furrow,
SI and S2 may be visible, and a suggestion in some individuals of S3. Genal region slopes gently downward,
outward and forward, to a narrow, wire-like border (PI. 1, fig. 1), which is best developed anteriorly, appearing
to die out laterally without reaching the genal angle. Posterior margin transverse proximally, distally curving
slightly back to an acute genal angle, not prolonged by a spine; no posterior border or border furrow, but
posterior edge upturned. Curved eye lobe close to anterior portion of glabella, posterior end in transverse line
with SI. Branches of facial suture directed outward and forward from eye lobe, anterior straight and at about
45° to sagittal line, posterior at about 70°, curved slightly back distally. No specimen shows evidence of a
doublure laterally, but occasional specimens (PI. 1, figs 4-5; PI. 2, fig. 4) show what appears to be a ventral
plate beneath the frontal area of the cranidium. The anterior margin of the plate coincides with the dorsal
cephalic margin, but has a groove along the margin (mould of a ventrally projecting ridge). The lateral margin
of the plate appears to be a straight line, lying beneath the course of the anterior branch of the suture. The
posterior margin coincides medially with the anterior margin of the glabella, laterally it is directed out to a
position outside the anterior end of the eye lobe. In the figured specimens the plate lies at a level below that
of the free cheeks; this and the distinct posterior and lateral edges suggest that this structure is a ventral plate
and not part of the crumpled dorsal exoskeleton. It is regarded as an unusual rostral plate, not separated
laterally from the doublure of the cheek by a connective suture, there being no sign of the presence of such a
doublure.

Thorax of 14 segments (PI. 1, fig. 1 ; PI. 2, figs 1, 3), in occasional specimens 15 (PI. 1, fig. 4). Axial region
tapers gradually behind sixth ring; rings with transverse posterior margin, rarely (PI. 1, fig. 4; PI. 2, fig. 1) the
forwardly-curved arc impressed into the ring indicates the presence of the doublure. Axial furrow shallow.

WHITTINGTON: BURLINGIID TRILOBITES

II

ill-defined. In relatively broader specimens (PI. 2, fig 1) the inner half of first pleura is directed transversely,
outer portion curving gently back to falcate tip. The inner, transverse portion of the pleura is reduced in width
(tr.) in successive pleurae and is gone by about the eighth pleura. In relatively narrower specimens (PI. 2,
fig. 3) there is no transverse inner portion of the pleura. The outer portions of the pleurae curve progressively
more strongly posteriorly, and the 14th pair are directed exsagittally beside the pygidium. The tips of the
pleurae are backwardly pointed. Anterior edge of each pleura a raised ridge proximally, which fades away
distally. Posterior edge of each pleura bent up to overlie ridge of succeeding pleura. Between these transverse
marginal ridges the faintly concave pleura is not impressed by a pleural furrow. No specimen shows any
evidence of a pleural doublure, the tip being ill-defined.

Pygidium (PI. 1, fig. 4) narrow (tr.), parallel-sided, axial region not convex, faint parallel lines suggesting a
division into two pairs of backwardly directed pleurae, the inner pair shorter so that the posterior outline is
deeply notched. The posterior thoracic pleurae extend well beyond the pygidial pleurae.

Discussion. All the specimens from Canada are internal moulds, flattened or crumpled, except for
one incomplete example which appears to show the original convexity (PI. 1, figs 2-3). Reference
has been made to relatively broader and narrower specimens (PI. 2, figs 1, 3), probably the result
of tectonic distortion. The restoration (Text-fig. 1) is a compromise between these two types of
flattened specimens, and the single example showing the convexity. The crumpling, and the
indefinite ragged margins of the exoskeleton laterally, suggest that the cuticle was thin. In no
specimen is there any breakage at the lateral margin of cephalon or thorax to reveal the mould of
a doublure. The presence of the apparently ventral and relatively large rostral plate, beneath the
frontal area of the cranidium, is therefore most unusual. Its seemingly definite shape and posterior
margin suggest that it is real, and if the cuticle of the plate was thin, and lay close to the dorsal
exoskeleton, it was presumably rarely exposed. Two specimens from Canada (PI. 1, figs 1, 4) show
a slightly more convex median portion of the axial region, about half the width of the axis. It
extends almost the length of the thorax in the lectotype, being evident only posteriorly in the other
specimen. I suggest that this convexity may be the impression in the thin axial exoskeleton of a
partial filling of the alimentary canal.

The originals of Westergaard’s (1936) Bwlingia laevis (PI. 2, fig. 2; PI. 3, fig. 1) are also early
Middle Cambrian in age, from the Eccaparadoxides oelandicus zone ( E . pinus subzone) of Oland,
Sweden. They are differently preserved, and the holotype is about twice the size of all but one of
the Canadian specimens. It is an external mould in fine-grained shale, and has the outline of the
hypostome preserved on the glabella by a narrow, low ridge. The lateral margin is a curve concave
abaxially, the posterior portion wider, semicircular in outline; the anterior margin is not outlined,
and the sagittal line of the hypostome is only slightly displaced from that of the dorsal exoskeleton.
Thus the hypostome appears to be almost in position beneath the anterior portion of the glabella,
but no rostral plate is visible, only what are taken to be the anterior branches of the suture crossing
the frontal area of the cephalon. The right free cheek is displaced inwards, the left apparently in
place. Hence whether this hypostome was natant, or whether it was conterminant, i.e. attached by
a hypostomal suture to the rostral plate seen in the Canadian specimens, is uncertain. The second
specimen (PI. 3, fig. 1) in shale has a thin, light coloured layer adhering to it, presumably a
replacement of the original exoskeleton. The right genal region has the lateral border extending to
the genal angle; the latter appears acute because the posterior margin is broken and incomplete.
Both examples have 14 thoracic segments, the holotype showing the low axis of the pygidium
divided by 4 inter-ring furrows, the post-axial region ill-defined posteriorly. One additional
specimen (RS Ar 46253) from the same locality, counterpart moulds of a telescoped exoskeleton, is
smaller than the holotype and shows the notched posterior margin of the pygidium.

A small block of limestone was collected by Mr M. J. Collins in 1968 from Kloxasen, Jamtland,
Sweden (stop 9 of Thorslund and Jaanusson 1960, p. 47). It contains five specimens of Bwlingia
(those referred to by Bergstrom 1973, p. 18), a cranidium of Eccaparadoxides cf. oelandicus , and
Ellipsocephalus polytomus , and is of early Middle Cambrian age. In two of the specimens of
Bwlingia , the exoskeleton is exposed from the dorsal side, in the other three from the ventral. The
two best (PI. 3, figs 6-7) are of similar size to the Canadian specimens, and have 14 thoracic

12

PALAEONTOLOGY, VOLUME 37

segments and the narrow, elongate pygidium with the deep posterior notch. Some of the original
convexity is retained, and the overlap of the posterior edge of each pleura (and the cephalon), upon
the ridge of the following pleura is clearly shown. Some displacement and telescoping of segments
has occurred, and the ragged edges of the thin thoracic exoskeleton are well displayed. The rostral
plate may be pressed against the dorsal exoskeleton in the specimen exposed from the ventral side,
but the evidence is equivocal. The axial furrow is shallow, with no posterior pit in the thoracic
segments suggestive of an axial articulating process and socket, nor any evidence of fulcrum, fulcral
articulating devices, or of doublure in the thorax or pygidium.

Westergaard (1936, pi. 32) remarked on the lack of glabellar furrows, the outline of the thorax,
and the larger size as distinctive of his species, but acknowledged the close similarity to B. hectori.
Comparisons between specimens of similar size show no distinctive differences. Some Canadian
specimens (e.g. PI. 2, figs 1, 3) show little or nothing of glabellar furrows. The appearance of there
being a genal spine on the right side of the original of Plate 3, figure 1 , is in part because the posterior
edge of the right cheek is irregularly broken, the first thoracic segment pushed beneath it. Other
specimens (PI. 3, figs 6-7) show only a small point at the genal angle. I regard the Canadian and
Swedish specimens as probably representing one species.

The species Bur/ingia obscura Soloviev, 1969, was described from incomplete, crumpled and
partly disarticulated specimens in oil shale, from the northern region of the Siberian platform,
Anabar river basin, Amgan Stage, Kuonamsk horizon, Oryctocephalops frischenfeldi zone, of
early Middle Cambrian age. Soloviev’s reconstruction ( 1969, pi. 2, fig. 3) of an exoskeleton c. 8 mm
in length shows 12 thoracic segments, the last pair of pleurae directed outward at about 35° to the
sagittal line, beside a bilobed pygidium. The latter has a short (sag.) undivided axis, and the pleural
region lacks furrows. Soloviev questioned the number of thoracic segments, but regarded the
pygidium as characteristic of his species. One photograph (Soloviev 1969, pi. 1, fig. 10), however,
shows the posterior portion of the thorax with pleurae directed progressively more strongly
backward, and almost exsagittally posteriorly, much as in B. hectori. In the diagnosis of Burlingia ,
I have cited the characters of the better-known type species, rather than those of the apparently
distinctive Siberian species.

A single example of B. ovcita Zhou and Yuan, in Chang et al. 1980 (p. 380, fig. 105; pi. 110, fig. 1)
is from the early Middle Cambrian Kaili Formation, eastern Guizhou province, southwest China
(Chang 1988, p. 55). It appears similar to B. hectori , and is smaller than the Canadian examples.

Genus schmalenseeia Moberg, 1903

Type species. By monotypy ; S. amphionura Moberg, 1903, from the earliest Upper Cambrian, zone of Agnostus
pisiformis of Sweden.

Diagnosis. Glabella with SO, Sl-3 developed as moderately deep lateral depressions; in the Swedish
species thorax of 7 or 8 segments, pygidium with 7 or 8 pairs of fused pleurae, posterior pair directed
back behind axis.

Distribution. Early Middle Cambrian in Sweden, in the late Middle Cambrian of eastern Guizhou, China, and
Tasmania, Australia, and in the early Upper Cambrian of Newfoundland, England, Sweden and Siberia.

Schmalenseeia amphionura Moberg, 1903
Plate 3, figures 2-5

Diagnosis. Glabella tapering forward to rounded, blunt, frontal glabellar lobe, sagittal ridge on
preglabellar field, prominent boss adjacent to axial furrow, beside frontal glabellar lobe, branches
of facial suture lie on sutural ridge. Median node on occipital ring, each of the seven thoracic
segments, and the anterior axial rings of the pygidium.

WHITTINGTON: BURLING1ID TRILOBITES

13

Discussion. The original type material from the early Upper Cambrian Agnostus pisiformis zone of
Oland, Sweden, and flattened and crumpled specimens from the same horizon in Newfoundland,
have been described (Whittington 1981). Additional pieces of dark grey, fine-grained limestone from
the type locality in Oland in the SGU collections contain many fragmentary cranidia and pygidia.
The external mould of a small exoskeleton, original of Westergaard (1922, p. 1 19), from Skogsby,
Oland, has the thorax partly telescoped and the pygidium broken sagittally, but shows the original
convexity. The axial furrow is not impressed, but is a change in slope, the pleural region inclined
downward and outward, more steeply distally (not flattened distally as my 1981, fig. 1, restoration
implies). The irregularly broken posterior edge of cephalon and segments is upturned to overlie the
short (sag.), convex articulating half ring and ridge along the anterior margin of each pleura. A low,
narrow ridge appears to be present along the outer edge of each pleura; there is no evidence of any
doublure. Between the marginal ridges the pleura is flat, lacking a pleural furrow.

Both internal and external mould are present in the Newfoundland material, and one of the
former (PI. 3, fig. 5) shows a ventral plate in front of the glabella. The lateral margin of the plate
appears to coincide with the course of the anterior branch of the facial suture, and the median
portion of the posterior margin with the preglabellar furrow. Distally this posterior margin is
directed outward and backward to a point below the boss in the fixed cheek, and appears to lie
below it in the internal mould. On this slim evidence I suggest that a rostral plate like that in
Burlingia may have been present in Schmalenseeia.

Specimens of Schmalenseeia which resemble S. amphionura in having the median preglabellar
ridge, the boss on the fixed cheek beside the frontal glabellar lobe, well-defined glabellar furrows,
and in the form of the thoracic segments and pygidium, have been described from the early Upper
Cambrian of northern Siberia ( S . spinulosa Lazarenko, 1960) and central England (Rushton 1978).
Similar specimens of uppermost Middle Cambrian age include those named S. gostinensis Jago,
1972, from Tasmania, S. sinensis Yang, 1978, an exoskeleton from eastern Guizhou, and two species
based on isolated cranidida and pygidium from eastern China (Ju, in Qui et al., 1983). Relationship
between these forms is indicated by the characters they have in common, distinctions between them
being based on minor characters such as axial tubercles or spines on the glabella (Rushton 1978,
p. 274), or the continuation of the preglabellar ridge on to the frontal glabellar lobe (Jago 1972,
p. 233). Better material is needed to clarify specific distinctions.

Schmalenseeia acutangula Westergaard, 1948
Plate 4, figures 1-6

Holotype. SGU 6354, incomplete exoskeleton, by original designation of Westergaard 1948, pi. 1, fig. 2, from
a block of dark grey limestone, Gislovshammar, south-east coast of Scania, Sweden.

Other material SGU 6353, incomplete exoskeleton, original of Westergaard 1948, pi. 1, fig. 3; SGU 6355,
thorax and pygidium, original of same, pi. 1 , fig. 4, locality and horizon as holotype. SGU, further unnumbered
pieces from Gislovshammar, and one from a boring in Jamtland, Sweden.

Horizon. Tomagnostus fissus and Acidusus atavus zone, Paradoxides paradoxissimus beds, early Middle
Cambrian.

Diagnosis. Differs from S. amphionura in lacking sagittal preglabellar ridge and boss beside frontal
glabellar lobe, outer portion of occipital furrow and SI -3 less deeply impressed; eight thoracic
segments, no median axial node, interpleural ribs of pygidium less prominent posteriorly.

Discussion. In outline and convexity of the exoskeleton this species is similar to S. amphionura. The
narrow, raised cephalic border is present on the frontal area, and may be present laterally. The
upturned palpebral lobe is close to the glabella, and extends from opposite the frontal glabellar lobe
to opposite L3 ; the course of the facial sutures similar in the two species, but there is no evidence

14

PALAEONTOLOGY, VOLUME 37

of a sutural ridge in S. acutangula. The genal angle appears to be almost a right angle, and is not
prolonged backward in a blunt point or spine. Thoracic segments are like those of S. amphionura,
and show the short (sag. ), raised articulating half ring, the anterior marginal ridge of the pleura, and
upturned posterior edge. The thin exoskeleton is broken and irregular along this posterior edge, the
distal edge also broken and shows no evidence of a marginal ridge, nor of a facet. In the original
of Plate 4, figures 4 and 5, there appears to be an eighth pair of pleurae in the pygidium, faintly
defined, behind the tip of the axis. These pleurae appear to have been the shortest, the seventh, sixth
and fifth extended successively farther backward, so that the outline of the pygidium was indented
by a notch behind the axis.

Schmalenseeia jagoi sp. nov.

Plate 4, figures 7-10

Ho/otype. SGU 6357, internal mould of incomplete cranidium, original of Westergaard 1948, pi. 1, fig. 6, from
a block of dark limestone, Brantevik, south-east coast of Scania, Sweden.

Other material. SGU 6356, internal mould of cranidium, original of Westergaard 1948, pi. 1, fig. 5, same
locality as holotype. Unnumbered pieces of similar limestone from Gislovshammar, south-east coast of Scania,
Sweden, containing fragmentary internal moulds of cranidia.

Horizon. Tomagnostus fissus and Acidusus atavus zones, Paradoxides paradoxissimus beds, early Middle
Cambrian.

Diagnosis. Cranidium with facial sutures of Schmalenseeia type, differs from S. acutangula in that
glabella at base is one-third (rather than one-fifth) width (tr.) of cranidium at posterior margin;
palpebral lobe adjacent to axial furrow and longer (exs.), posterior end opposite L2 and
consequently postocular area of cheek relatively shorter (exs.); prominent median occipital node.
In larger cranidia outer end of occipital furrow and S 1—3 each a shallow, subcircular pit inside axial
furrow.

Discussion. Jago (1972, p. 233) first recognized the distinctive characters of this cranidium, and the
new species is named in his honour. Additional material from the same locality and horizon as those
figured, includes two external and five internal moulds of cranidia, poorly preserved and crumpled.
Only the largest of these shows glabellar furrows like those in the holotype, other lacking such
depressions; all show the median occipital tubercle. In the original of Plate 4, figs 8 and 10, and a
smaller, unfigured cranidium, the posterior margin appears complete, and the genal angle scarcely
prolonged. The posterior margin of the cheek in the holotype (PI. 4, figs 7, 9) is ragged and broken,
and the appearance of an acutely angulate genal angle on the right side may be because of this
breakage. Westergaard (1948, p. 4) regarded an acute genal angle as characteristic of his species,
presumably because of this specimen, but I omit this character as of doubtful validity.

The glabellar furrows of the larger cranidium, position of the eye lobe, course of the sutures, and
lack of the posterior border, combine to suggest retaining this species in Schmalenseeia , although
the thorax and pygidium are unknown. Jago’s view that this species, and acutangula , be removed
from Schmalenseeia is not accepted.

Acknowledgements. I am indebted to the following for the loan of specimens (abbreviations used in the text for
each institution are given in parentheses): Mr D. M. Rudkin, Royal Ontario Museum (ROM), Toronto,
Canada; Dr V. Berg-Madsen, Geological Survey of Sweden (SGU), Uppsala, Sweden; Mr F. J. Collier,
National Museum of Natural History (USNM), Washington, D. C., USA.; Professor J. Bergstrom,
Riksmuseum (RS), Stockholm, Sweden; Dr D. B. Norman, Sedgwick Museum (SM), Cambridge; Mr M. J.
Collins, Standish, Lancashire, UK. Dr A. W. A. Rushton kindly drew my attention to the occurrence of
Burlingia in Siberia. Mrs Sandra J. Last prepared the text. Miss Hilary Alberti drew Text-fig. 1, and Mr
B. K. Harvey made the negatives. I am grateful to the Department of Earth Sciences, Cambridge University,

WHITTINGTON: BURLINGIID TRILOBITES

15

for the use of facilities, and to the Leverhulme Foundation for the support of my research. This is Cambridge
Earth Sciences Publication no. 3080.

REFERENCES

beecher, c. E. 1897. Outline of a natural classification of the trilobites. American Journal of Science , Series 4 ,
3, 89-106, 181-207.

Bergstrom, j. 1973. Organisation, life and systematics of trilobites. Fossils and Strata , 2, 1-69, pis 1-6.
chang, wen-tang 1988. The Cambrian System in Eastern Asia, correlation chart and explanatory notes.
International Union of Geological Sciences , Publication , 24, 1-81, 4 charts.

— (ZHANG WENTANG), ZHANG SENGUI, LIN HUANLING. ZHU ZHAOLING, QIAN YIYUAN, ZHOU ZHIYI, YUAN

jinliang and lu yanhao. 1980. Cambrian trilobite faunas of southwestern China. Palaeontologica Sinica
( new series B), 16, 1—497, pis 1-134. [In Chinese].

fortey, r. a. 1990. Ontogeny, hypostome attachment and trilobite classification. Palaeontology , 33, 529-576.

— and chatterton, b. d. e. 1988. Classification of the trilobite suborder Asaphina. Palaeontology, 31,
165-222.

— and owens, R. m. 1990. Trilobites. 121-142. In McNamara, k. j. (ed.). Evolutionary trends. Belhaven Press,
London, xviii + 368 pp.

fritz, w. H. 1971. Geological setting of the Burgess Shale. Proceedings of the North American Paleontological
Convention , September 1969 , Part /, Extraordinary Fossils , Allen Press, Lawrence, Kansas, 1155-1170.

— 1990. Comments: in defense of the escarpment near the Burgess Shale fossil locality. Geoscience Canada ,

17, 106-110.

henningsmoen, g. 1951. Remarks on the classification of trilobites. Norsk geologisk Tidsskrift , 29, 174-217.
jago, j. b. 1972. Two new Cambrian trilobites from Tasmania. Palaeontology , 15, 226-237.

LAZARENKO, N. P. 1960. In KRYSKOV, L. N., LAZARENKO, N. P., OGIENKO, L. V., and CHERNYSHEVA, N. E. 21 1-255,
pis 50-53, in markovsy, b. p. (ed.) [New species of prehistoric plants and invertebrates of the U.S.S.R., part
2] VSEGEI, Moscow. [In Russian],

moberg, J. c. 1903. Schmalenseeia amphionura , en ny trilobit-type. Geologiska Foreningens i Stockholm
Forhandlingar , 25, 93-102.

moore, r. c. (ed.). 1959. Treatise on invertebrate paleontology. Part O. Arthropoda I . Geological Society of
America and University of Kansas Press, Boulder, Colorado and Lawrence, Kansas, xix + 560 pp.

QIU, H. A., LU, Y. H., ZHU, Z. L., BI, D. C., LIN, T. R., ZHOU, Z. R., ZHANG, Q. Z., QIAN, Y. Y., JU, T. Y., HAN, N. R.
and wei, x. 1983 [Palaeontological atlas of East China, I, Early Palaeozoic]. Geological Publishing House,
Beijing, ix + 657 pp, 176 pis. [In Chinese].

rasetti, f. 1951. Middle Cambrian stratigraphy and faunas of the Canadian Rocky Mountains. Smithsonian
Miscellaneous Collections , 116, 1-277, pis 1-34.

reed, F. R. c. 1899. Woodwardian Museum Notes: a new trilobite from Mount Stephen, Field, B.C. Geological
Magazine, New Series, Decade 4, 6, 358-361.

robison, r. a. 1984. Cambrian Agnostida of North American and Greenland. Part 1, Ptychagnostidae.
University of Kansas, Paleontological Contributions, 109, 1-59.

— 1988. Trilobites of the Horn Dal Formation (late Middle Cambrian), central North Greenland.
Meddelelser om Gronland, Geoscience , 20, 23-103.

rushton, a. w. a. 1978. Fossils from the Middle-Upper Cambrian Transition in the Nuneaton District.
Palaeontology, 21, 245-283.

scotese, c. r. and mckerrow, s. w. 1990. Revised World maps and introduction. 1-21. In mckerrow, w. s.and
SCOTESE, c. R. (eds). Palaeozoic palaeography and biogeography. Memoirs of the Geological Society, London,
12, 435 pp.

shergold, J. h. 1969. Oryctocephalidae (Trilobita: Middle Cambrian) of Australia. Bulletin of the Bureau of
Mineral Resources, Geology and Geophysics, Australia, 104, 1-66, pis 1-12.

Soloviev, i. a. 1969. [Discovery of representatives of the genus Burlingia Walcott in the Amgan Stage of
Northern Siberia]. Uchenye Zapiski nauclmo-issledovatef skikh Instituta Geolgiska Arktika, 26, 9-12, pis 1-2.

[In Russian].

Stubblefield, c. J. 1936. Cephalic sutures and their bearing on current classifications of trilobites. Biological
Reviews, 1 1 , 407^140.

thorslund, p. and jaanusson, v. 1960. The Cambrian, Ordovician and Silurian in Vastergotland, Niirke,
Dalarna and Jemtland, central Sweden. International Geological Congress, XXIst Session, Norden I960,
Excursions A 23 and C18, 51 pp.

16

PALAEONTOLOGY, VOLUME 37

walcott, c. D. 1908. Cambrian trilobites. Smithsonian Miscellaneous Collections , 53, 13-52.
westergaard, a. H. 1922. Sveriges Olenidskiffer. Sveriges Geologiska Undersokning, Series C, 18, 1-205,
pis 1-16.

— 1936. Paradoxides oelandicus beds of Oland. Sveriges Geologiska Undersokning , Series C, 394, 1-66,
pis 1-12.

— 1948. Non-agnostidean trilobites of the Middle Cambrian of Sweden. Sveriges Geologiska Undersokning ,
Ser. C, 498, 1-32, pis 1-4.

Whittington, H. B. 1957. The ontogeny of trilobites. Biological Reviews, 32, 421-469.

— 1981. Paedomorphosis and cryptogenesis in trilobites. Geological Magazine, 119, 591-602.

— 1990. Articulation and exuviation in Cambrian trilobites. Philosophical Transactions of the Royal Society
of London, Series B, 329, 27—46.

yang jia-lu. 1978. [Middle and Upper Cambrian trilobites of western Hunan and eastern Guizhou].
Professional Papers of Stratigraphy and Palaeontology, 4, 1-82, pis 1-13. Geological Publishing House,
Peking, China. [In Chinese].

H. B. WHITTINGTON
Sedgwick Museum
Department of Earth Sciences

Typescript received 3 September 1992 University of Cambridge

Revised typescript received 8 March 1993 Cambridge CB2 3EQ, UK

TINY PLAGIAULACOID MULTITUBERCULATE
MAMMALS FROM THE PURBECK LIMESTONE
FORMATION OF DORSET, ENGLAND

by Z. KIELAN-JAWOROWSK A and P. C. ENSOM

Abstract. Albionbaatar gen. nov., assigned to the Albionbaataridae fam. nov., in the Plagiaulacoidea, is
described upon the basis of A. denisae sp. nov. based on minute (about 1-5 mm long) upper premolars (P5) and
three anterior upper premolars (0-5 — 0-9 mm long), all from the Purbeck Limestone Formation of southern
England. Two incomplete lower premolars (p4), each found in a sample yielding an upper premolar of
Albionbaatar , are described as Plagiaulacoidea, fam., gen. et sp. indet., and we speculate whether they may be
counterparts of Albionbaatar. Albionbaatar differs from known multituberculates in having upper premolars
with greater numbers of very small cusps, arranged in three longitudinal rows, and P5 with extensive lingual
slope covered by transverse ridges. The oblique lingual slope on P5, characteristic of Albionbaatar , occurs in
previously known plagiaulacoids, where there are no ridges. The structure of Albionbaatar shows that the
correlation between tooth size and number of cusps characteristic for the Cimolodonta (at least for the
Taeniolabidoidea) may not hold for all the Plagiaulacoidea.

Since 1986, one of us (P. C.E.) has been washing and screening vertebrate-rich clay deposits of the
Purbeck Limestone Formation, principally from Sunnydown Farm Quarry near Langton
Matravers, Swanage, Dorset, southern England. The residues have yielded fish, amphibians, reptiles
and mammals (Ensom 1987, 1988; Ensom et al. 1991). The mammalian collection contains
members of the Triconodonta, Docodonta, Symmetrodonta, Peramura and Multituberculata.
Previously we described the collection of multituberculate teeth recovered from this site prior to
1988 (Kielan-Jaworowska and Ensom 1992). We showed that the multituberculates from the
Purbeck Limestone Formation contain, in addition to the Plagiaulacoidea, members of the
Paulchoffatoidea, hitherto unrecorded in the Purbeck Limestone Formation. That material allowed
us to propose a new subordinal classification of Late Jurassic and Early Cretaceous multi-
tuberculates (Kielan-Jaworowska and Ensom 1991, 1992), and furnished new data on enamel
microstructure in plagiaulacoid multituberculates (Fosse et al. 1991).

In our earlier paper on Purbeck multituberculates from the Purbeck Limestone Formation of
Dorset (Kielan-Jaworowska and Ensom 1992) we omitted reference to one complete upper
premolar (DORCM GS 23) and one fragmentary lower premolar (DORCM GS 22), the affinities
of which were in contention. Now we feel sufficiently confident of their multituberculate identity to
include them in this paper. Additional multituberculate specimens have been recovered from the
residues. This new material contains four, possibly conspecihc, upper teeth similar to DORCM
GS 23, and identified by us as the upper premolars, and a fragmentary lower premolar which was
found in a sample yielding an upper premolar. The upper premolars differ from those occurring in
the known multituberculates and we assign them to Albionbaatar denisae gen. et sp. nov., which we
place in the family Albionbaataridae nov., in Plagiaulacoidea.

For the terminology of the teeth described, localities and sampling methods, see Kielan-
Jaworowska and Ensom (1992). The seven teeth described in this paper derive from five samples
(Nos: 40-14-1 kg; 70 - 26-5 kg; 77 - 29-2 kg; 78 - 34 0 kg; 95 - 38-0 kg), all collected from the
Cherty Freshwater Member, Purbeck Limestone Formation, near Langton Matravers, Purbeck,
Dorset.

I Palaeontology, Vol. 37, Part 1, 1994, pp. 17-31, 2 pls.|

© The Palaeontological Association

18

PALAEONTOLOGY, VOLUME 37

We employ the following abbreviations: BMNH - Natural History Museum (previously British
Museum (Natural History)), London; DORCM - Dorset County Museum, Dorchester, Dorset,
England; YPM - Peabody Museum, Yale University, New Haven, Connecticut, USA.

MULTITUBERCULATE CUSP FORMULAE

Krause (1977) argued that the cusp formulae should reflect the homologies of cusp rows. He stated
(Krause 1977, p. 5): ‘In most ptilodontoid and taeniolabidoid species there are three rows of cusps
on P4; external, middle and internal. The number of cusps in each row is given as a formula x:y:z,
where x, y, and z designate the number of cusps in the external, middle and internal rows
respectively. The middle row (y) assuming that it is homologous in all ptilodontoid and
taeniolabidoids, is the longest and most cuspidate.’ However, Krause did not present evidence to
show that the middle (and not for example, internal) row of cusps is homologous in all
ptilodontoids and taeniolabidoids. Assuming that the middle row is homologous in all the
Cimolodonta, he further argued that the internal row of cusps is absent in several ptilodontoid
genera, and the apparently neomorphic ‘row’ of cusps is often developed labial to the external row
(x) of cusps, on an anteroexternal bulge. Consequently he suggested the cusp formula in P4 of
Ptilodus montanus should not be listed in a traditional way as 0-3:5-8:9-10, but rather as
(0-3)5-8:9-10:0. If Krause’s proposal were applied to Albionbaatar denisae , described below, the
cusp formula for P5 should not be written as 5-7: 7:7 (as we do), but rather as (5-7)7: 7:0, since
the buccal row in the P5 of Albionbaatar may not be homologous with the buccal row of other
plagiaulacoid P5s. We continue to list cusp formulae in all the upper premolars of Albionbaatar in
the traditional way, as we are of the opinion that with the present limited state of our knowledge
on the homology of the cusp rows, it is impossible to demonstrate unequivocally which cusp rows
are homologous within the various multituberculate groups. In addition, we are concerned that
changes to the terminology may lead to confusion, because the same element would be differently
named in new papers.

STRATIGRAPHY

The teeth described in this paper were extracted from the top 20-40 mm of a clay within the Cherty Freshwater
Member of the Purbeck Limestone Formation, about 2-6 m below the base of the Cinder Member. The clay
and overlying micritic limestone can be correlated with beds DB 102 and 103 respectively, of the Durlston Bay
Purbeck Limestone Formation stratotype (Clements 1969, 1993). West (1988) described both the clay and
overlying limestone at Sunnydown Farm Quarry and concluded that they were deposited in environments with
‘very low’ and ‘low’ salinities respectively. West suggested that the clays represented extensive mud flats
bordering a freshwater lake. The presence of dinosaurs is evident from the considerable number of footprints
recorded (Ensom 1987, 1988).

In our earlier discussion (Kielan-Jaworowska and Ensom 1992) of the stratigraphy of the Sunnydown Farm
Quarry site and other localities, we pointed out that the Cherty Freshwater Member was currently taken as
uppermost Jurassic, with the overlying Cinder Member marking the base of the Cretaceous. This was in line
with the published accounts of Rawson et al. (1978), following Casey (1963). Consequently, we had described
the mammalian fauna as Upper Jurassic (Tithonian), but Lower Cretaceous (Berriasian) may be more
appropriate. After our paper was accepted for publication, a review and revision of the correlation of the
Purbeck-Wealden in north-west Europe (Allen and Wimbledon 1991) appeared. This highlighted the
continuing debate over the position of the Jurassic-Cretaceous boundary in southern England, and the Wessex
sub-basin in which these mammal-bearing strata lie was discussed. Allen and Wimbledon pointed out that
palynomorph and ostracode evidence would place the Jurassic-Cretaceous boundary more or less at the base
of the Purbeck Limestone Formation if the Berriasian Berriasella jacobi subzone is ratified as the basal subzone
of the Cretaceous. They also noted that use of the marine Subthurmannia subalpina subzone as the basal
Berriasian subzone would place the Jurassic-Cretaceous boundary at or slightly below the Cinder Member,
more or less in line with Casey’s proposal. They concluded that an international consensus is required on this
matter before the final decision is made. This uncertainty over the age of the material should be born in mind
when considering this fauna within the framework of mammalian evolution.

KIELAN-JAWOROWSKA AND ENSOM: PURBECK MAMMALS

19

SYSTEMATIC PALAEONTOLOGY

Suborder plagiaulacoidea (Simpson, 1925) Hahn, 1969
Family albionbaataridae fam. nov.

Diagnosis. As for the only known genus Albionbaatar gen. nov.

Stratigraphical and geographical range. Known only from the Late Jurassic or Early Cretaceous Purbeck
Limestone Formation of Southern England.

Genera. The family is monotypic.

Genus albionbaatar gen. nov.

Derivation of name. Albion - the oldest, possibly Celtic, name of England, baatar - Mongolian, a hero, a suffix
often used for the names of multituberculate genera.

Diagnosis. Shrew-sized multituberculate; differs from all multituberculates in having relatively flat,
multicusped anterior upper premolars, with ten to fourteen cusps arranged in three rows, rather
than three or four, rarely up to nine high cusps in two rows, and in having lingual slope in all
premolars covered by prominent, subparallel ridges. Differs from Paulchoffatoidea, known
Plagiaulacoidea, Taeniolabidoidea and non-specialized Ptilodontoidea in having P5 with three rows
of numerous small cusps. Differs from Paulchoffatoidea in having P5 distinctly longer relative to
width. Differs from known Plagiaulacoidea in lack of dramatic difference in the size of the cusps in
the same upper premolar. Shares with few derived Cimolodonta (e.g. Ptilodus) numerous cusps on
posterior upper premolars.

Albionbaatar denisae sp. nov.

Plate I ; Plate 2, figures 1-5, 9-10

Derivation of name. In honour of Dr Denise Sigogneau- Russell, in recognition of her work on Mesozoic
mammals.

Holotype. DORCM GS 212 (sample 40) right P5 in a fragment of a maxilla (PI. 1, figs 3-6; PI. 2, fig. 10).

Type horizon and locality. Cherty Freshwater Member, Purbeck Limestone Formation; near Langton
Matravers, Purbeck, Dorset.

Other material. DORCM GS 227 (sample 77) right P5; and tentatively assigned: DORCM GS 24 (sample 70)
?right ?P3; DORCM GS 282 (sample 78) ?left ?P2; DORCM GS 23 (sample 95); ?left ?P1, all from the type
horizon and locality.

Diagnosis. P5 roughly 8 shaped, with buccal and lingual margins incurved in the middle; the
holotype specimen is 1-4 mm long, 0 85 mm wide. Cusp formula 5-7: 7: 7. Buccal row shorter than
lingual and median ones, with cusps decreasing in size posteriorly. Cuspules or crenulations present
buccal to buccal row.

Description. The holotype specimen, right P5 (DORCM GS 212; PI. 1, figs 3-6; PI. 2, fig. 10), consists of the
crown with two roots, preserved in a small piece of a maxilla. This P5 is roughly 8-shaped in occlusal view,
having buccal and lingual margins indented in the middle. It is slightly asymmetrical in this view, as the lingual
slope is a little higher (especially posteriorly) than the buccal one. Both anterior and posterior margins are
gently rounded, the posterior being somewhat pointed in the middle. The anterior wall is concave in the middle,
possibly to receive the posterior part of the crown of the preceding tooth.

There are seven cusps in the lingual row and the radiating ridges of these cusps continue onto the lingual
slope (PI. 1, figs 3-4; PI. 2, fig. 10). There are two or three ridges, that start at the tip of each cusp and continue
upwards, diverging at the beginning, and then continuing roughly subparallel, some bifurcating in the middle.

20

PALAEONTOLOGY, VOLUME 37

all bifurcating on the dorsal part of the slope, the dorsal margin of which is slightly convex and rounded. The
lingual ridges of the first three cusps are not abraded and extend upwards and slightly anteriorly along the
whole height of the slope. In the midlength of the upper half of the slope, there extends a very weak,
longitudinal furrow, about 0-60 mm long and 0T0 mm wide, situated 0T8 mm below the middle part of the
dorsal margin of the tooth and 0 33 mm below the posterior part of the margin. The lingual ridges of the fourth
and fifth cusps are broken by this furrow and continue upwards above it, reaching the lingual margin, but are
less distinct than in the ventral part of the slope. The furrow becomes very shallow posteriorly, extending along
the anterior part of the wear facet described below, and disappears 0 26 mm in front of the posterior margin.
The lingual ridges of the last two cusps are almost completely worn away. There is a flat wear facet that
occupies the posterior quarter of the surface of the lingual slope and reaches the base of the crown. The arrow
in Plate 2, figure 10 denotes the direction of poorly preserved wear striations. The height of the lingual slope
in the holotype specimen is 0-80 mm anteriorly, 0-66 mm in the middle and 0-83 mm posteriorly.

There are seven cusps in the median row. The buccal and occlusal views appear to show a median row count
of eight, as the crenulated anterior margin looks like an additional cusp. The four first are of equal size, the
fifth is the smallest, the sixth larger again, the seventh the largest of all. The difference in cusp size is not great.
Posteriorly and slightly lingually of the seventh cusp, there is a cuspule with four radiating ridges. The cusps
of the buccal row are less regular than those of the lingual and median rows and vary more strongly in size;
defining them as cusps, cuspules or crenulations is sometimes difficult. We recognize five cusps in the anterior
half of the tooth, the first of which is lower than the others, ridge-like and elongated longitudinally. The second
and third cusps are similar in size and shape to those of the lingual and medial rows, the third being the largest
of all in the buccal row. The fourth and fifth cusps are smaller than the third, situated close to each other and
joined by a longitudinal ridge running across their tips. To the rear of the fifth cusp, there are irregular
crenulations, posterior to which three cuspules extend to the end of the tooth. Buccal to the buccal row there
is a single extra cusp, situated opposite the level between the second and the third cusps of the buccal row. To
the rear of the extra cusp, the buccal margin, as seen in occlusal view, is developed as an elevated ridge. In
buccal view (PI. 1, fig. 5) this ridge can be seen to be crenulated, which gives an appearance of the presence
of the fourth row of cusps, which is not the case. In buccal view, the short vertical ridges extend from this ridge
vertically to the buccal margin of the tooth; on the anterobuccal slope in front of the extra cusp, similar ridges
are more prominent.

The anterior root, exposed in buccal view, is 1 mm long; only the base of the posterior root is visible (in
posterior view).

A right P5 (DORCM GS 227; PL 1, figs 7-8; PI. 2, fig. 9) has been preserved as two fragments that have
been glued together. Because of the small size of the tooth, it has proved difficult to reassemble it in its original
anatomical position; the posterior region of the tooth should be arranged more horizontally and the posterior
end turned more lingually than has been done; after being glued the tooth is possibly longer than it originally
was. The bases of both roots are preserved. The tooth is 0 93 mm wide; its length cannot be established,
because of the break. It differs from DORCM GS 212 in being less symmetrical, having its anterior margin
protruding more strongly in its lingual part than buccally, the posterior margin being less rounded, and the
indentations of the buccal and lingual margins situated more anteriorly. The cusps in DORCM GS 227 are
insignificantly larger (up to 0T8 mm in diameter rather than up to 0T5 mm in DORCM GS 212). In DORCM
GS 227 there are six fully developed cusps in the lingual row and a cusp-like crenulation on the anterior margin
giving an appearance of one more cusp. In the median row there are also six fully developed cusps and an
anterior cusp-like crenulation of the anterior margin. The important difference with respect to DORCM GS 212
concerns the cusps of the buccal row, which decrease in size posteriorly in both specimens. In DORCM GS 227,
there are seven cusps (the third one broken), which up to the end of the row are developed as fully

EXPLANATION OF PLATE 1

Figs 1-8. Albionbaatar denisae gen. et sp. nov. Cherty Freshwater Member, Purbeck Limestone Formation;
Sunnydown Farm Quarry near Langton Matravers, Dorset, England. 1-2, DORCM GS 24; ?right ?P3;
occlusal and oblique medio-occlusal views. 3-6, DORCM GS 212, holotype; right P5 with a fragment of a
maxilla ; occluso-hngual, lingual, buccal and occlusal views. 7-8, DORCM GS 227, plastic cast ; broken right
P5, glued from two parts; occlusal and buccal views.

All are scanning electron micrographs (all except figure 4 are stereopairs). All teeth are oriented with anterior
margin up, and are x 30.

PLATE 1

KIELAN-JAWOROWSKA and ENSOM, Albionbaatar

22

PALAEONTOLOGY, VOLUME 37

recognizable cusps. In DORCM GS 212 only five fully developed cusps are recognized, and the posterior part
of the row is formed by crenulations and cuspules. The single extra cusp present in DORCM GS 212 buccal
to the buccal row at the level between the second and third cusps is less obvious in GS 227, resembling rather
a crenulation.

The height of the lingual slope in DORCM GS 227 is 0 80 mm above the third cusp, and 0-83 mm above the
fifth cusp. The pattern of ridges on the lingual slope differs slightly, but not significantly from DORCM GS
212. The longitudinal furrow extends roughly along the mid-height of the lingual slope and is placed more
posteriorly (above the posterior half of the fourth cusp and last three cusps) than in DORCM GS 212. As in
the latter the anterior part of the lingual slope is not abraded. The posterior part, opposite the posterior half
of the fourth cusp, and the fifth to seventh cusps, is abraded and the degree of wear increases posteriorly. Above
the posterior parts of the fourth and fifth cusps and just below the longitudinal furrow, there are small wear
facets; the ridges above the furrow are not abraded. Above the sixth cusp there is a large, rectangular wear facet
that extends between the tip of the cusp and the longitudinal furrow ; only the ridges above the furrow and the
first lingual ridge of the sixth cusp are not abraded. Above the seventh cusp the whole area between the tip of
the cusp and the longitudinal furrow is completely abraded producing a rectangular wear facet which fits
lightly to the longitudinal furrow. Above the furrow the ridges of the seventh cusp are abraded producing a
slightly concave wear facet. The seventh cusp is also abraded posteriorly and the crescent-shaped wear facet
is present between its tip and the posterior margin of the tooth.

We regard the above cited differences between the two P5s as due to the individual variation within a single
species (see Discussion).

A ?left ?P1 (DORCM GS 23; PI. 2, figs 1-2) is 0-9 mm long lingually, 0-76 mm buccally and 0-73 mm wide,
roughly rectangular in occlusal view. The posterior margin protrudes posteriorly in the lingual part. The cusp
formula is 3:5:4 with one anterior and three buccal cuspules. The cusps of the lingual slope increase in size
posteriorly. The median row has an irregular course; the first three cusps increase in size posteriorly and are
arranged longitudinally. The fourth cusp, of approximately the same size as the third, is displaced lingually.
The fifth cusp, the smallest, is displaced buccally. There is a small cuspule in front of the buccal row and three
cusps in this row which increase in size posteriorly. Buccal to the buccal row there are three cuspules, the first
two situated opposite the posterior parts of the first and second cusps, the last one at the posterobuccal corner
of the third cusp. All the cuspules are covered with ridges similar to those of the cusps. The lingual slope is
covered by the ridges of the lingual cusps, which disappear before reaching the tooth-margin, which is partly
broken in this specimen. There are three to four ridges diverging (some bifurcating) from the tip of each cusp
up the lingual slope. It cannot be discounted that the tooth is a right one.

A ?left ?P2 (DORCM GS 282; PI. 2, figs 3-5) is suboval, almost rectangular, wider posteriorly than
anteriorly, 0 8 mm long and 0-56 mm wide (measured at one-third of the length from the posterior end). It is
possible that the tooth was slightly wider as the buccal margin is partly damaged. The posterior margin is

EXPLANATION OF PLATE 2

Multituberculates from the Cherty Freshwater Member, Purbeck Limestone Formation; Sunnydown Farm
Quarry, near Langton Matravers, Dorset, England.

Figs 1-2. Albionbaatar denisae sp. nov. DORCM GS 23; ?left ?P1, roots not preserved; occlusal and dorsal
views, 1, x 30; 2, x 20.

Figs 3-5. Albionbaatar denisae sp. nov. DORCM GS 282; ?left ?P2; lingual, occluso-lingual and occlusal views;
x 30.

Figs 6-7. Plagiaulacoidea, farm, gen. et sp. indet. DORCM GS 22; posterior part of left p4, lingual side
broken; buccal and lingual views. 6, x30; 7, x 20.

Figs 8, 1 1. Plagiaulacoidea, farm, gen. et sp. indet. DORCM GS 25, plastic cast; anterior part of p4; occlusal
and buccal views; x 30.

Fig. 9. Albionbaatar denisae sp. nov. DORCM GS 227, plastic cast; right P5; oblique linguo-occlusal view;
x 30.

Fig. 10. Albionbaatar denisae sp. nov. DORCM GS 212, holotype; right P5; lingual view, only lingual slope
shown; arrow denotes the direction of poorly preserved wear striations; x 70.

All are scanning electron micrographs; the specimen in figures 3-5 was not coated with gold; figures 1-8 are
stereo-pairs. Upper teeth in figures 1-3 are oriented with anterior margin up, in figures 4-5 and figures 8-9
anterior is to the left.

PLATE 2

KIELAN-JAWOROWSKA and ENSOM, Albionbaatar , Plagiaulacoidea

24

PALAEONTOLOGY, VOLUME 37

incurved opposite the buccal row and therefore the tooth is shorter buccally than lingually. The cusp formula
is 4 : 5 : 4. The anterior cusps in the median and buccal rows (buccal broken) are slightly smaller than the others.
The lingual row is relatively high, covered with the ridges of the lingual cusps that extend along its height and
disappear before reaching the margin. The posterobuccal part of the slope is smooth, but no wear facet is
discernible. It cannot be discounted that the tooth is a right one. It is only slightly shorter than DORCM
GS 23, but narrower in respect to the length.

A ?right ?P3 (DORCM GS 24; PI. 1, figs 1-2) is roughly trapezoid, with the buccal margin shorter than the
lingual; 0-7 mm long lingually, 0-5 mm long buccally and 0-6 mm wide. The anterior margin is gently rounded;
the posterior margin projects posteriorly opposite the lingual row of cusps. The cusp formula is 4:4:4. The
anterior and posterior cusps in the buccal row and the posterior cusp in the median row are smaller than other
cusps. The lingual slope is covered with the ridges of the lingual cusps. Two, three or four ridges extend dorsally
from each cusp, diverging from the tips; some bifurcate and disappear before reaching the margin of the tooth.
The ridges of the two middle cusps are more prominent than the others. The lingual slope is higher than the
buccal, especially anteriorly, and is not abraded. It cannot be discounted that the tooth is a left one. It is the
smallest tooth of the whole series.

DISCUSSION

The five teeth assigned to Albionbaatar denisae gen. et sp. nov. are identified as upper premolars,
because they bear the conical cusps, covered by radiating ridges, characteristic of multituberculate
premolars (Krause et al. 1992).

We assign the two P5s (DORCM GS 212 and GS 227) to the same species, although they differ
slightly in size, as well as in details of the arrangement and size of the cusps, because the amount
of variation in size, cusp number, and structure exhibited by these teeth falls within the range of
variation seen in larger samples of multituberculates that have been assigned to a single species (see
e.g. Clemens 1963a; Krause 1977, 19826, 1982c, 1987; Rigby 1980; Fox 1980; Johnston and Fox
1984; and many others).

We regard the five upper premolars described above as conspecific, as they display numerous
small cusps of the same type arranged in three rows, and similar type of ridges that extend onto the
lingual slope. In addition they all roughly fit each other in size.

We identify DORCM GS 212 and DORCM GS 227 as P5s for the following reasons. In
plagiaulacoid multituberculates there are five upper premolars. P4 and P5 have been preserved
together only in Bolodon osborni Simpson, 1928 (BMNH 47735a); Ctenacodon laticeps (Marsh,
1881) (YPM 11761); Psalodon potens (Marsh, 1887) (YPM 11834), and in the as yet undescribed
plagiaulacoid skull from the Morrison Formation, not available to us for this study (Simpson 1928,
1929; Kielan-Jaworowska et al. 1987; Engelmann et al. 1990; Kielan-Jaworowska and Ensom
1992). In the first three taxa mentioned above there is a continuous extensive lingual wear facet on
P4, directed anteroposteriorly and slightly obliquely linguodorsally-buccoventrally. Wear on P5 is
more extensive than on P4 and consists of two facets, the anterior one extending for about two-
thirds of the tooth length, roughly parallel to the wear facet on P4, and the posterior one arranged
obliquely anterolingually-posterobuccally with respect to the sagittal plane (Kielan-Jaworowska
and Krause in preparation). The wear pattern of DORCM GS 212 and DORCM GS 227 is different
from that on P4 in known Plagiaulacoidea, and is reminiscent of that on P5. However, unlike other
plagiaulacoids, in both studied specimens the anterior part of the lingual slope is not worn, and only
on the posterior part is a distinct wear facet present. The lack of wear on the anterior part may be
due to the young age of the individuals at the time of death. One can presume that in older
individuals the wear facet may be developed also on the anterior part of the slope, as characteristic
of the Plagiaulacoidea. P5 in the Plagiaulacoidea displays the difference in wear of the anterior and
posterior parts of the tooth, while the ultimate upper premolars, P4s in Ptilodontoidea (Krause
1982a; see also Wall and Krause 1992) and in eucosmodontid Taeniolabidoidea (personal
observations of Z.K.-J.) are worn only in the posterolingual part. The wear striations on the wear
facet in both studied P5s are poorly preserved; among the two they are better seen in DORCM
GS 212 (PI. 2, fig. 10) and seem to follow roughly the same (posterodorsal) direction as in P4 of
Ptilodus (Krause 1982a, fig. 9), possibly homologous to P5 in Plagiaulacoidea (Clemens 1963a).

KIELAN-JAWOROWSKA AND ENSOM: PURBECK MAMMALS

25

Another reason for identifying these teeth as P5s is their elongate shape. The length/ width ratio
in DORCM GS 2 1 2 is 1 -64 : 1 , in GS 227 the estimated ratio is about 116:1. Known Plagiaulacoidea
show the same variation in relative lengths of P4 and P5. In the Allodontidae, both P4 and P5 are
relatively short in relation to the width (Simpson 1929; Kielan-Jaworowska et al. 1987). However,
in the Plagiaulacidae, e.g. Bolodon osborni , the length/width ratio of P5 is 1-6:1, with an extensive
lingual slope, and in P4 the length/width ratio is 1-39:1 with a low lingual slope (Simpson 1928;
Kielan-Jaworowska et al. 1987; Kielan-Jaworowska and Ensom 1992). P4 and P5 are unknown in
other taxa (Clemens 1963a; Clemens and Lees 1971 ; Kielan-Jaworowska et al. 1987; Bakker and
Carpenter 1990; Engelmann et al. 1990).

In the Cimolodonta (except the specialized Taeniolabididae) P4 (an apparent homologue of P5
in the Plagiaulacoidea, see above) is greatly elongated, although the length/width ratio in various
cimolodonts may vary considerably, and the premolars in front of it are all very short. That is why
we identify DORCM GS 212 and 227 as P5s rather than P4s. We are confident that DORCM
GS 212 and DORCM GS 227 are right teeth, because of the presence of the above mentioned wear
facet at only one end (evidently posterior) of the lingual slope.

The identifications of the three anterior upper premolars as PI, P2 or P3 can be only very
tentative. Because of the lack of wear facets on these teeth, we were unable to identify them as right
or left teeth with any certainty. We therefore put two question marks at each identification, one
concerning the side and the other the number of the tooth. We identify DORCM GS 23 as a ?left
?P1, as the cusps on PI are larger than on successive premolars in a plagiaulacine Bolodonl
elongatus. Also, in Bolodon osborni and Ctenacodon laticeps , PI is larger and has slightly larger cusps
than P2 and P3. However, this rule does not always hold, as for example, in Bolodon crassidens , PI
is smaller than P2 and P3 (Simpson 1928, 1929; Kielan-Jaworowska et al. 1987; Kielan-
Jaworowska and Ensom 1992). We identify DORCM GS 24 as the ?right ?P3, by comparison with
Bolodon osborni , Bolodonl elongatus , Ctenacodon laticeps and Arginbaatar dimitrievae (Kielan-
Jaworowska et al. 1987) in which P3 is the smallest of all the upper teeth. We tentatively identify
DORCM GS 282 as ?left ?P2, as it is intermediate in size between DORCM GS 23 and DORCM
GS 24.

plagiaulacoidea, fam., gen. et sp. indet.

Plate 2, figures 6-7.

Material. A fragment of the posterior part of the buccal wall of a left p4 (DORCM GS 22, sample 95).

Description. We figure and discuss DORCM GS 22 in this paper, as it was found in the sample yielding
DORCM GS 23, identified by us as ?P1 of Albionbaatar denisae. Teeth found in the same sample may belong
to the same individual (Kielan-Jaworowska and Ensom 1992).

The preserved fragment of the posterior part of the buccal wall is L3 mm long along the dorsal margin and
about 1-1 mm high. The preserved fragment shows that the tooth was roughly rectangular and longer than
high. The height of the tooth increases anteriorly and the dorsal margin is almost straight. Along the dorsal
margin there are seven projections (the seventh incomplete), which are provided with very short anteroventrally
directed ridges, more prominent on the two complete anterior projections than posteriorly. The most anterior
ridge is 0-2 mm long. The ridge of the last projection is directed more vertically downwards than the anterior
ones and widens at the end. In front of the two last ridges there are small rounded depressions, positioned
anteroventrally to the serrations between the projections. The posterior margin is vertical in the upper part and
bulges posteriorly opposite the posterobuccal cusps, situated in the ventral part of the crown. The
posterobuccal cusps are strongly worn from the dorsal side and their joint dorsal margin forms a prominent
shelf, concave upwards. Due to the wear it is difficult to estimate how many posterobuccal cusps were originally
present; traces of at least five cusps are recognizable. The length of the row of posterobuccal cusps was about
0-6 mm. The preserved part shows that the posterobuccal cusps were prominent, protruding buccally over the
surface of the tooth more strongly than in other plagiaulacoid mammals. Along the lower margin of the tooth
is a longitudinal furrow. When observed in dorsal view, p4 appears narrower (despite the lingual wall having
been broken) than in other multituberculates. Rather than flaring laterally, the buccal wall is arranged almost
straight downwards in respect to the narrow dorsal margin, which is indicative of a very narrow tooth. This

26

PALAEONTOLOGY, VOLUME 37

is probably the reason why the lingual wall had been broken, a state of preservation rarely found in
multituberculate p4s. Another unusual feature of this tooth is that the ventral margin of the buccal wall is bent
inwards and the buccal wall possibly overhung the roots, of which there are no traces.

Remarks. The roughly rectangular shape of DORCM GS 22 is indicative of the Plagiaulacoidea.
The tooth differs from the known plagiaulacoid and other multituberculate p4s in being more knife-
like, with buccal and lingual walls placed closer together, rather than flaring outward. In addition,
the posterobuccal cusps protrude buccally to a greater extent than in hitherto recorded
plagiaulacoids. There are shorter ridges and a ventral horizontal furrow, and the buccal wall of the
crown is bent inwards. These two latter features do not occur, as far as we are aware, in known
multituberculate taxa. It is impossible to conclude whether DORCM GS 22 is a counterpart of the
upper premolars of Albionbaatar denisae. Given the structure of the upper premolars of
Albionbaatar not known to occur in other multituberculates, one can presume, albeit inconclusively,
that the p4 in Albionbaatar might have a somewhat different structure from those in plagiaulacoid
multituberculates, e.g. similar to that in DORCM GS 22.

plagiaulacoidea, fam. gen. et sp. indet.

Plate 2, figures 8, 1 1

Material DORCM GS 25 (sample 70), anterior part of the left p4.

Description and comments. We figure and discuss DORCM GS 25 in this paper, as it was found in the sample
yielding DORCM GS 24, identified as ?P3 of Albionbaatar denisae.

The size of DORCM GS 25 is difficult to estimate because of its incompleteness. DORCM GS 25 is
reminiscent of plagiaulacine genera from Purbeck strata of Great Britain (Simpson 1928; Kielan-Jaworowska
et al. 1987; Kielan-Jaworowska and Ensom 1992) in having a gently vaulted dorsal margin, with distinct
projections, but it differs from that of Plagiaulax becklesii and IPlioprion falconeri in being much smaller, and
in having relatively longer ridges that are subhorizontally oriented. It is reminiscent of p4 of Bolodon osborni
and B. minor in having the distance between the first and second projections distinctly longer than between the
succeeding ones, but differs from them in having longer ridges which are subhorizontally orientated and in
having different lengths between the dorsal projections. In DORCM GS 25 the first projection is situated
0-46 mm to the rear of the anterior margin; the distance between the first and second is 0-5 mm, between the
second and third 0-36 mm, and between the third and fourth 0-21 mm. In both species of Bolodon the distance
between the first and second projections is longer than between the next ones (as in DORCM GS 25), but the
distances between the successive ones are of subequal lengths. The anterior wall of DORCM GS 25 (PI. 2, fig. 8),
although damaged, is similar to B. osborni in having a triangular dorsal plate and crenulated margins
(Kielan-Jaworowska and Nessov 1992, fig. 3A).

DORCM GS 25 is similar to p4s in plagiaulacine taxa. However, given the strong heterodonty between
upper and lower premolars in multituberculates, it cannot be ruled out that it might be a counterpart of
Albionbaatar denisae.

ALLOCATION OL ALBIONBAATAR WITHIN MULTITUBERCULATA

Albaionbaatar cannot be assigned to the Haramiyoidea (Hahn 1973; Sigogneau- Russell 1989) as all
the cusps on the premolars are of equal height and covered by radiating ridges, rather than the cusps
being of different heights and lacking ridges. The upper premolars of the Theroteinidae, assigned
to Allotheria (Sigogneau- Russell et al. 1986; Hahn et al. 1989) are not known.

Albionbaatar cannot be assigned to the Paulchoffatoidea (Hahn 1969, 1971, 1977; Krause and
Hahn 1990), as its fifth upper premolar (P5) is much more elongated in relation to its width than
in the Paulchoffatoidea.

Plagiaulacoidea, which are common in the Purbeck Limestone Formation of England (Simpson
1928; Kielan-Jaworowska et al. 1987; Kielan-Jaworowska and Ensom 1992), have an extensive
lingual slope in P5 as in Albionbaatar , but in contrast to Albionbaatar , this slope is not covered by

KIELAN-JAWOROWSKA AND ENSOM: PURBECK MAMMALS

27

ridges. The P5 of Albionbaatar differs from most (but not all) known members of the Plagiaulacoidea
in being more elongated in relation to its width. The most important difference is the presence of three
rows of very small and low cusps, which do not differ dramatically in size from each other, rather
than one and a half or two rows of relatively large and high cusps, strongly varying in size. Another
important difference concerns the number of cusps. In P5 of Albionbaatar denisae there are nineteen
cusps and six cuspules in DORCM GS 212, and twenty-one cusps and four cuspules in DORCM
GS 227. This differs dramatically from the known Plagiaulacoidea, where the number of cusps
varies in P5 from six to seven in the Allodontidae, seven cusps and eight cuspules in Bolodon osborni,
five cusps and seven cuspules in Bolodon cf. minor , and six cusps in Eobaatar magnus. In P4 there
are seven cusps in the Allodontidae, seven cusps and four cuspules in Bolodon osborni , two cusps
and four cuspules in Bolodon crassidens and eight cusps and two cuspules in Eobaatar magnus
(Owen 1854, 1871; Falconer 1857, 1862; Marsh 1887; Simpson 1928, 1929; Kielan-Jaworowska et
al. 1987; Kielan-Jaworowska and Ensom 1992).

Neither can Albionbaatar be assigned to the Cimolodonta (Ptilodontoidea and Taeniolabidoidea)
(Granger and Simpson 1929; Simpson 1937; Sloan and Van Valen 1965 ; Kielan-Jaworowska 1970,
1974; McKenna 1975; Krause 1977, 1982a, 19826; Clemens and Kielan-Jaworowska 1979;
Archibald 1982; Krause and Carlson 1987), as P5 has an extensive lingual slope covered by
subparallel ridges, which is not the case in P4 (homologous to P5 of Plagiaulacoidea - Clemens
1963a) of the Cimolodonta. In the taeniolabidid Taeniolabidoidea and in the Cimolomyidae
(Granger and Simpson 1929; Clemens 1963a; Sloan and Van Valen 1965; Clemens and Kielan-
Jaworowska 1979; Archibald 1982), P4 tends to be reduced and this group does not invite a
comparison with Albionbaatar. In eucosmodontid taeniolabidoids and in most ptilodontoids
(Granger and Simpson 1929; Kielan-Jaworowska 1970, 1974; Clemens and Kielan-Jaworowska
1979) P4 has only one complete (lingual) row of cusps and a short anterobuccal one. However, in
some advanced ptilodontoids (e.g. Ptilodus), P4 has two more or less complete rows of cusps and
a short anterobuccal one, sometimes without cusps. The numbers of cusps in P4s in Ptilodus and
Prochetodon may reach twenty and are thus comparable to those in P5 of Albionbaatar (Granger
and Simpson 1929; Simpson 1937; Krause 1977, 1982u, 1982 b, 1987). However, the ptilodontoid
P4 differs dramatically from P5 of Albionbaatar in the lack of an extensive lingual slope covered with
transverse ridges.

Still more striking differences concern the anterior upper premolars. In Albionbaatar these teeth
are relatively flat, and bear ten to fourteen small, equal-sized cusps arranged in three rows, whereas
in all other multituberculates these teeth bear two to nine high cusps (most often three or four)
arranged in two rows.

The upper premolars of the specialized Early Cretaceous Asian Arginbaataridae bear a small
number of cusps and differ markedly from those of Albionbaatar (Kielan-Jaworowska et al. 1987).
Another specialized multituberculate family, the South American Ferugliotheriidae (Bonaparte
1986; Krause et al. 1992), is known from isolated molars and tentatively assigned anterior upper
premolars. These latter teeth do not show a multicusped structure characteristic of Albionbaatar.

It follows from the foregoing comparisons that Albionbaatar , because of the presence of an
extensive lingual slope in P5, is best attributed to the Plagiaulacoidea, although it differs
considerably from the known members of this suborder.

Cole and Krause (1988, p. 12a) studied the relationships between tooth size and cusp
numbers in Cimolodonta, and concluded: 'These relationships suggest that, at least for taenio-
labidoids, tooth size and cusp numbers are not independent and, therefore, that their covariance
should be addressed in future studies of multituberculate systematics.’ The present study shows that
in Plagiaulacoidea as a whole there is no correlation between tooth size and number of cusps, but
it seems possible that such a correlation might be found within some plagiaulacoid lines.
Albionbaatar is amongst the smallest known multituberculates, with minute teeth, which at the same
time are characterized by very numerous cusps. Numerous cusps on the upper premolars, but
correlated with a large size, were until now known only in advanced representatives of the
Cimolodonta.

28

PALAEONTOLOGY, VOLUME 37

CONCLUDING REMARKS

Multituberculates were found in the Purbeck Limestone Formation of England as early as the
middle of the nineteenth century (Falconer 1857, 1862). In this paper we have described several
minute upper premolars from the rocks of this formation, assigned to the Albionbaataridae fam.
nov. within Plagiaulacoidea, and two lower premolars which might be their counterparts.

These teeth have been recovered using the washing and screening technique widely used in
vertebrate palaeontology since Hibbard’s paper (1949). Palaeontologists searching for Mesozoic
mammals often use this technique with sieve meshes down to 1 mm (personal observation). This
material has been extracted using a bulk sieving machine constructed to the design of Ward (1981)
with sieves of mesh aperture size 0-33 mm. The teeth described here have at least one dimension
below 1 mm. Similar teeth may be present in other mammal-bearing strata but previously have been
missed.

Because of their superficial similarity with rodents, multituberculates have been sometimes
referred to as the rodents of the Mesozoic. Their ecology has been discussed by palaeontologists
since the middle of the past century. It has been generally accepted that these were the first mammals
that occupied herbivorous niches, although Krause (1982a) demonstrated that they might have
been omnivorous. Krause and Jenkins (1983) maintained that some Palaeocene North American
multituberculates might have been scansorial in their life-style. Miao (1988) and Kielan-Jaworowska
and Qi (1990) suggested fossorial adaptation of Palaeocene-Eocene genus Lambdopsalis. Kielan-
Jaworowska and Gambaryan are presently describing a collection of the postcranial skeletons of
Late Cretaceous multituberculates from the Gobi Desert and showed that multituberculates had
abducted limbs. They suggested on the basis of both anatomical and sedimentological evidence, that
the life-styles of these multituberculates were similar to those of the modern nocturnal murid
rodents (gerbils) that live in semi-desert habitats. However, because of the different limb position,
multituberculate locomotion was possibly different from those of small modern rodents, as well as
of other small therian mammals which have parasagittal limbs.

The members of the Albionbaataridae belong not only to the smallest multituberculates but
possibly to the smallest known mammals as well. As they are known only from a few isolated teeth,
the reconstruction of their mode of life poses difficulties.

The Purbeck Limestone Formation is a sequence of limestones, clays and shales of marginal and
non-marine origin, at times showing strong terrestrial influence. Within the sequence there is clear
evidence of both periods of aridity with evaporite formation, and of freshwater influx permitting the
growth of charophytes accompanied by an abundance of freshwater ostracodes and gastropods.
Certain horizons within the Cherty Freshwater Member have evidence of both sets of conditions
with halite pseudomorphs, charophytes and freshwater invertebrates in close association (Ensom
1985), probably representing seasonal variations of climate. The presence of amphibians in the
fauna at Sunnydown Farm Quarry does point to the existence of wetter conditions, in keeping with
West’s (1988) shallow lake, and certainly more humid than that which existed in the Late
Cretaceous of the Gobi Desert.

Multituberculates range in time from the Late Triassic to the Early Oligocene (Hahn and Hahn
1983) and are thus the longest lived order of mammals. It is to be expected that during their long
evolutionary history they would adapt themselves to different ecological niches and display an array
of life-styles.

Acknowledgements. We are very grateful to Professor David W. Krause and Dr Denise Sigogneau-Russell for
reading the draft of this manuscript and providing most useful comments, and to Dr R. G. Clements for
commenting on our discussion on the stratigraphy. We thank the Dorset Natural History and Archaeological
Society for permitting us to work on material from the excavation which they funded with the permission of
Mr and Mrs R. F. Notley. Acknowledgement was made in our previous account to the many individuals who
assisted with the excavation and subsequent processing of samples. However, inadvertently we omitted to

KIELAN-JAWOROWSKA AND ENSOM: PURBECK MAMMALS

29

thank the Community Enterprise Scheme based at the Dorset County Council’s Durlston Country Park,
Swanage, Dorset, which contributed in a variety of ways to the success of the project. Most of the scanning
electron micrographs were taken by Z. K.-J. and Ms T. Rolfsen (Institute of Biology, University of Oslo), the
remainder by Z. K.-J. with the help of Dr L. A. Nessov and Mr J. Hurum at the Paleontological Museum,
University of Oslo. The research of Z. K.-J. was supported by funds from the Norges Allmennvitenskapelige
Forskningsrad (Grant ABC/LR 441.92/003).

REFERENCES

allen, p. and Wimbledon, w. a. 1991. Correlation of NW European Purbeck Wealden (non-marine Lower
Cretaceous) as seen from the English type-areas. Cretaceous Research , 12, 511-526.

Archibald, j. d. 1982. A study of Mammalia and geology across the Cretaceous-Tertiary boundary in Garfield
County, Montana. University of California Publications in Geological Sciences , 122, i— xvi -I- 1-286.
barker, r. t. and carpenter, k. 1990. A new latest Jurassic vertebrate fauna, from the highest levels of the
Morrison Formation at Como Bluff, Wyoming. Part III. The mammals: a new multituberculate and a new
paurodont. Hunteria , 2, 4—8.

bonaparte, j. f. 1986. Sobre Mesungulatum houssayi y nuevos mamiferos Cretacicos de Patagonia, Argentina.

Actas IV Congreso Argentino de Paleontologia , 2, 48-61.

Casey, R. 1963. The dawn of the Cretaceous Period in Britain. Bulletin of the South-Eastern Union of Scientific
Societies, 117, 1-15.

clemens, w. a. 1963 a. Fossil mammals of the type Lance Formation, Wyoming. Part I. Introduction and
Multituberculata. University of California Publications in Geological Sciences, 48, i-vi+ 1-102.

19636. Wealden mammalian fossils. Palaeontology, 6, 55-69.

— and kielan-jaworowska, z. 1979. Multituberculata. 99-149. In lillegraven, j. a., kielan-jaworowska,
z. and clemens, w. a. (eds). Mesozoic mammals: the first two-thirds of mammalian history. University of
California Press, Berkeley, Los Angeles, London, x + 311 pp.

— and lees, p. m. 1971. A review of English Cretaceous mammals. In kermack, d. m. and kermack, k. a.
(eds). Early mammals. Zoological Journal of the Linnean Society, 50, Supplement 1, 117-130.

Clements, R. G. c. 1969. Annotated cumulative section of the Purbeck Beds between Peverill Point and the
Zig-Zag Path, Durlston Bay. Fig. A 35a-35d. In torrens, h. s. (ed. ). International Field Symposium on the
British Jurassic, Excursion No. 1, Guide for Dorset and South Somerset. A1-A71.

— 1993. Type section of the Purbeck Limestone Group, Durlston Bay, Swanage, Dorset. Proceedings of the
Dorset Natural History and Archaeological Society, 114, 181-205.

cole, t. m. and krause, d. w. 1988. Interspecific relationships between tooth size and cusp numbers in the
Multituberculata (Mammalia). Journal of Vertebrate Paleontology, 8, Supplement to Number 3, 12a.
engelmann, G. f., greenwald, n. s., callison, G. and chure, d. j. 1990. Cranial and dental morphology of
a Late Jurassic multituberculate mammal from the Morrison Formation. Abstract. Journal of Vertebrate
Paleontology, 10, Supplement to Number 3, 22 a.

ensom, p. c. 1985. An annotated section of the Purbeck Limestone Formation at Worbarrow Tout, Dorset.
Proceedings of the Dorset Natural History and Archaeological Society, 106, 87-91.

— 1987. A remarkable new vertebrate site in the Purbeck Limestone Formation. Proceedings of the Dorset
Natural History and Archaeological Society, 108, 205-206.

1988. Excavations at Sunnydown Farm, Langton Matravers, Dorset: amphibians discovered in the
Purbeck Limestone Formation. Proceedings of the Dorset Natural History and Archaeological Society, 109,
148-150.

— evans, s. e. and milner, a. r. 1991. Amphibians and reptiles from the Purbeck Limestone Formation
(Upper Jurassic) of Dorset. 19-20. In kielan-jaworowska, z., heintz, n. and nakrem, h. a. (eds). Fifth
Symposium on Mesozoic Terrestrial Ecosystems and Biota: Extended Abstracts. Contributions from the
Paleontological Museum, University of Oslo, 364, 70 pp.

falconer, h. 1857. Description of two species of fossil mammalian genus Plagiaulax from Purbeck. Quarterly
Journal of the Geological Society, London, 13, 261-282.

1862. On the disputed affinity of the mammalian genus Plagiaulax from the Purbeck Beds. Quarterly
Journal of the Geological Society, London, 18, 348-369.
fosse, g., kielan-jaworowska, z. and ensom, p. c. 1991. Enamel ultrastructure of Late Jurassic
multituberculate mammals. 25-26. In kielan-jaworowska, z., heintz, n. and nakrem, h. a. (eds). Fifth
Symposium on Mesozoic Terrestrial Ecosystems and Biota: Extended Abstracts. Contributions from the
Paleontological Museum, University of Oslo, 364, 70 pp.

30

PALAEONTOLOGY, VOLUME 37

fox, R. c. 1980. Mammals from the Upper Cretaceous Oldman Formation, Alberta. IV. Meniscoessus Cope
(Multituberculata). Canadian Journal of Earth Sciences, 17, 1480-1488.
granger, w. and simpson, G. G. 1929. A revision of the Tertiary Multituberculata. Bulletin of the American
Museum of Natural History , 56, 601-676.

hahn, G. 1969. Beitrage zur Fauna der Grube Guimarota Nr. 3. Die Multituberculata. Palaeontographica,
Abteilung A, 133, 1-100.

— 1971. The dentition of the Paulchoffatiidae (Multituberculata, Upper Jurassic). Memorias Servigos
Geologicos, Portugal, 17, 7-39.

— 1973. Neue Zahne von Haramiyiden aus der Deutschen Ober-Trias und ihre Beziehungen zu den
Multituberculaten. Palaeontographica, Abteilung A, 142, 1-15.

— 1977. Neue Schadel-Reste von Multituberculaten (Mamm.) aus dem Malm Portugals. Geologica et
Palaeontologica, 11, 161-186.

— and hahn, r. 1983. Multituberculates. In westphal. f. (ed.). Fossilium Catalogus, I: Animalia, pars 127.
Kugler Publications, Amsterdam, 409 pp.

— sigogneau-russell, d. and wouters, G. 1989. New data on Theroteinidae - their relations with
Paulchoffatiidae and Haramiyidae. Geologica et Palaeontologica, 23, 205-215.

hibbard, c. w. 1949. Techniques of collecting microvertebrate fossils. Contributions from the Museum of
Paleontology, University of Michigan, 8, 7-19.

johnston, p. a. and fox, r. c. 1984. Paleocene and Late Cretaceous mammals from Saskatchewan, Canada.
Palaeontographica , Abteilung A, 86, 163-222.

kielan-jaworowska, z. 1970. New Upper Cretaceous multituberculate genera from Bayn Dzak, Gobi Desert.
Palaeontologia Polonica, 21, 35 49.

1974. Multituberculate succession in the Late Cretaceous of the Gobi Desert (Mongolia). Palaeontologia
Polonica, 30, 23 44.

— dashzeveg, d. and trofimov, b. a. 1987. Early Cretaceous multituberculates from Mongolia and a
comparison with Late Jurassic forms. Acta Palaeontologica Polonica, 32, 3-47.

— and ensom, p. c. 1991. Suprageneric taxa of Late Jurassic and Early Cretaceous multituberculate
mammals. 35-36. In kielan-jaworowska, z., heintz, n. and nakrem, h.-a. (eds). Fifth Symposium on
Mesozoic Terrestrial Ecosystems and Biota: Extended Abstracts. Contributions from the Palaeontological
Museum, University of Oslo, 364, 70 pp.

— 1992. Multituberculate mammals from the Upper Jurassic Purbeck Limestone Formation of
Southern England. Palaeontology, 35, 95-126.

— and nessov, l. a. 1992. Multituberculate mammals from the Cretaceous of Uzbekistan. Acta
Palaeontologica Polonica, 37, 1-17.

— and Qi, T. 1990. Fossorial adaptations of a taeniolabidoid multituberculate mammal from the Eocene of
China. Vertebrata PalAsiatica, 28, 81-94.

krause, d. w. 1977. Paleocene multituberculates (Mammalia) of the Roche Percee local fauna, Ravenscrag
Formation, Saskatchewan, Canada. Palaeontographica, Abteilung A, 159, 1-36.

— 1982a. Jaw movement, dental function, and diet in the Paleocene multituberculate Ptilodus. Paleobiology ,
8, 265-281.

I982A Multituberculates from the Wasatchian land-mammal age. Early Eocene, of Western North
America. Journal of Paleontology, 56, 271-294.

1982c. Evolutionary history of Early Cenozoic Multituberculata (Mammalia), with emphasis on the
Family Ptilodontidae. Unpublished Ph.D. thesis. University of Michigan, Ann Arbor.

— 1987. Systematic revision of the genus Prochetodon (Ptilodontidae, Multituberculata) from the Late
Paleocene and Early Eocene of western North America. Contributions from the Museum of Paleontology,
University of Michigan, 27, 221-234.

and carlson, s. J. 1987. Prismatic enamel in multituberculate mammals: tests of homology and polarity.
Journal of Mammalogy, 68, 755-765.

— and hahn, g. 1990. The systematic position of the Paulchoffatimae (Multituberculata, Mammalia).
Journal of Paleontology, 64, 1051-1054.

— and jenkins, f. a., jr. 1983. The postcranial skeleton of North American multituberculates. Bulletin of the
Museum of Comparative Zoology, Harvard, 150, 199-246.

-kielan-jaworowska, z. and bonaparte, j. f. 1992. Ferugliotherium Bonaparte, the first known
multituberculate from South America. Journal of Vertebrate Paleontology, 12, 351-376.
mckenna, m. c. 1975. Toward a phylogenetic classification of the Mammalia. 21^16. In luckett, w. p. and
szalay, F. s. (eds). Phylogeny of the primates. Plenum Press, New York, 483 pp.

KIELAN-JAWOROWSKA AND ENSOM: PURBECK MAMMALS

31

marsh, o. c. 1887. American Jurassic Mammals. American Journal of Science, 32, 327-348.
miao, d. 1988. Skull morphology of Lambdopsalis bulla (Mammalia, Multituberculata) and its implications to
mammalian evolution. Contributions to Geology, University of Wyoming, Special Paper , 4, i-viii+1 104.
owen, R. 1854. On some fossil reptilian and mammalian remains from the Purbecks. Quarterly Journal of the
Geological Society, London , 10, 420-433.

— 1871. The fossil Mammalia of the Mesozoic formations. Monograph of the Palaeontographical Society,
33, (110), i— vi-H 1—115, pis l^L

RAWSON, P. F., CURRY, D., DILLEY, F. C., HANCOCK, J. M., KENNEDY, J. W., NEALE, J. W., WOOD, C. J. and WORSSAM,

b. c. 1978. A correlation of Cretaceous rocks in the British Isles. Special Report of the Geological Society,
London, 9. 1-70.

rigby, j. k. 1980. Swain Quarry of the Fort Union Formation, Middle Paleocene (Torrejonian), Carbon
County, Wyoming: Geological setting and mammalian fauna. Evolutionary Monographs, 3, i-vi+1 179.
sigogneau-russell, d. 1989. Haramiyidae (Mammalia, Allotheria) en provenance du Trias superieur de
Lorraine (France). Palaeontographica, Abteilung A, 206, 137-198.

— frank, r. m. and hemmerle, j. 1986. A new family of mammals from the lower part of the French
Rhaetic. 99-108. In padian, k. (ed. ). The beginning of the age of dinosaurs. Cambridge University Press,
Cambridge, 378 pp.

simpson, G. G. 1928. A catalogue of the Mesozoic Mammalia in the Geological Department of the British Museum.
British Museum (Natural History), London, x + 215 pp.

— 1929. American Mesozoic Mammalia. Memoirs of the Peabody Museum of Yale University, 3,
i-xv+ 1-235.

1937. Skull structure of the Multituberculata. Bulletin of the American Museum of Natural History, 73,
727-763.

sloan, r. e. and van valen, l. 1965. Cretaceous mammals from Montana. Science, 148, 220-227 .
wall, c. e. and krause, d. w. 1992. A biomechanical analysis of the masticatory apparatus of Ptilodus
(Multituberculata). Journal of Vertebrate Paleontology , 12, 172-187.
ward, d. !. 1981. A simple machine for bulk processing of clays and silts. Tertiary Research, 3, 121-124.
west, i. m. 1988. Notes on some of the sediments associated with the dinosaur footprints at Sunnydown Farm,
near Langton Matravers, Dorset. Proceedings of the Dorset Natural History and Archaeological Society , 109,
153-154.

Z. KIELAN-JAWOROWSKA
Paleontologisk Museum
Sars gate 1, Universitetet i Oslo
N-0562 Oslo 5, Norway

p. c. ENSOM

Typescript receved 25 November 1992
Revised typescript received 2 April 1993

Yorkshire Museum
Museum Gardens
York YOl 2DR, UK

A NEW ANGU1MORPH LIZARD FROM THE
JURASSIC AND LOWER CRETACEOUS OF

ENGLAND

by SUSAN E. EVANS

Abstract. A new genus of anguimorph lizard Parviraptor (type species P. estesi sp. nov.) has been identified
from middle Jurassic (Bathonian), and late Jurassic (Tithonian) or early Cretaceous (Berriasian) sites in
England. The genus is also represented in the late Jurassic of Portugal (Oxfordian or Kimmeridgian,
Guimarota). Parviraptor differs markedly from the only previously recorded Jurassic anguimorph,
Dorsetisaurus , and shares a complex of characters which suggests that it was a stem-platynotan, and that the
'Platynota’ had differentiated by the middle Jurassic.

The published Jurassic record of crown-group lizards is drawn from a relatively small number of
late Jurassic localities of which Solnhofen (Kimmeridgian, Germany), Guimarota (Oxfordian or
Kimmeridgian, Portugal), and Como Bluff (Tithonian, North America) have been amongst the
most productive (Estes 1983). Further important lizard material has come from the Purbeck Beds
of Dorset, England. These beds were, until recently, considered to be of late Jurassic (Tithonian)
age, and their vertebrate assemblage closely resembles that of Como Bluff. Allen and Wimbledon
(1991), however, based on ostracode and palynomorph data, suggested that the Purbeck Limestone
Formation may be of earliest Cretaceous (Berriasian) age. Representatives of at least three of the
four currently recognized lizard subgroups - namely Gekkota, Scincomorpha and Anguimorpha,
have been reported from Jurassic localities.

Jurassic anguimorphs have, until now, been represented by a single genus, Dorsetisaurus , known
from Purbeck (Hoffstetter 1967), Como Bluff (Prothero and Estes 1980), and Portugal (the latter
mistakenly under the name Introrsisaurus , Seiffert 1973). The supposed anguimorph Lisboasaurus ,
also from Portugal (Seiffert 1973), is a small theropod (Milner and Evans 1991), possibly a bird.
Living anguimorphs are broadly divided into two groups - ‘Anguioidea’ (anguids, anniellids and
xenosaurs, but the monophyly of this group is weakly substantiated (Estes et al. 1988)) and
Varanoidea ( sensu Pregill et at. 1986, varanids, lanthanotids and helodermatids). Varanoids form
part of a larger group, ‘Platynota’ (sensu Pregill et al. 1986), which also encompasses a range of
extinct genera including dolichosaurs, mosasaurs and their probable relatives - aigialosaurs. The
taxonomic position of Dorsetisaurus in relation to the two main anguimorph lineages is uncertain,
but previous workers (e.g. Hoffstetter 1967; Estes 1983; Borsuk-Bialynicka 1984) have placed it
closer to anguioids. There is certainly nothing to suggest platynotan affinity and the earliest
undisputed platynotans are currently the middle Cretaceous (Cenomanian-Turonian) aigialosaurs
from Croatia (reviewed by Carroll and deBraga 1992). Kuhn (1958) described briefly a specimen
from Solnhofen which he interpreted as a primitive platynotan, and named Proaigialosaurus.
Unfortunately, the specimen was in an unnamed private collection and its present location is
unknown. Hoffstetter (1964) suggested that Proaigialosaurus was a juvenile of the sphenodontian
Pleurosaurus, but he had not examined the specimen. Kuhn’s (1958) sketch represents the teeth as
sharply pointed. If this is correct, then identity with Pleurosaurus , which has a typically
sphenodontian dentition, is unlikely. Without the specimen, however, only speculation is possible.

Work on a middle Jurassic (Bathonian) microvertebrate assemblage from Kirtlington Cement

| Palaeontology, Vol. 37, Part 1, 1994, pp. 33— 49|

© The Palaeontological Association

34

PALAEONTOLOGY, VOLUME 37

text-fig. 1. Parviraptor estesi gen. et sp. nov. BMNH 48388, holotype; Durlston Bay, Dorset; Purbeck

Limestone Formation; x2.

text-fig. 2. Parviraptor estesi gen. et sp. nov. BMNH 48388, holotype; Durlston Bay, Dorset; Purbeck
Limestone Formation. Key; r, right; 1, left; Mx, maxilla; O, osteoderm; Pa, parietal; PI, palatine;
Pt, pterygoid; Sq, squamosal; ?V, possible vertebral fragment. The arrow (top right) indicates the isolated
archosaur tooth mentioned in the text. Scale bar represents 5 mm.

Works Quarry in Oxfordshire has yielded a new anguimorph which is strikingly different from
Dorsetisawus and shows characters suggestive of platynotan affinity. This new form is represented
at Kirtlington by jaws, skull bones and vertebrae, associated on the basis of size, bone texture and,
for the skull bones, complementary facets. Isolated bones of the same genus have subsequently been
identified in material from Guimarota (Evans personal observations) but, until recently, there was

EVANS: MESOZOIC LIZARDS FROM ENGLAND

35

no associated material. However, two partly associated specimens have recently been identified
from Purbeck material in the collections of The Natural History Museum, London (BMNH). The
first is a small block (BMNH 48388) found amongst undescribed material and bearing an
association of skull bones, including a maxilla (Text-figs 1-2). The second (BMNHR8511)
(Text-figs 3-4) was recovered from the pterosaur collection. The two specimens complement each
other, and confirm the association of skull and vertebral elements suggested from Kirtlington. Since
BMNH 48388 combines jaw and skull material, it is here designated as the holotype.

SYSTEMATIC PALAEONTOLOGY

SQUAMATA Oppel, 1911
anguimorpha Fiirbringer, 1900
‘platynota’ (sensu Pregill et al. 1986)

Genus parviraptor gen. nov.

Derivation of name. From the Latin parvus meaning small, and raptor meaning robber.

Type species. Parviraptor estesi gen. et sp. nov.

Range. Middle Jurassic (Bathonian) to late Jurassic (Tithonian) or early Cretaceous (Berriasian) of England,
with further material from the late Jurassic (Oxfordian or Kinuneridgian) of Portugal.

Diagnosis. A small (snout-vent length c. 150 mm) lizard characterized by a long, low maxilla,
bearing at least twenty long, narrow and sharply recurved teeth with a circular cross-section; teeth
pleurodont with shallow implantation, but each tooth position flanked by a build up of attachment
bone; teeth attached to broad alveolar border; in lower jaw (not preserved on the holotype), deep
lingual shelf present but subdental ridge lost, small intramandibular septum with fused lower edge;
frontals and parietals paired; parietal foramen retained; well-developed subolfactory processes on
frontals, meeting or nearly meeting in the midline; dorsal roofing bones without attached sculpture
but compound sculptured osteoderms perhaps present on at least some parts of body; low maxillary
facial process suggesting depressed antorbital region ; elongated narial margin suggesting reduced
contact between maxilla and nasal; palatine short and wide, narrow vomerine process, reduced
pterygoid process; no palatine teeth and no choanal groove; prefrontal/palatine contact weak or
ligamentous; pterygopalatine joint apparently weak; pterygoid long and slender, sharply incised by
suborbital fenestra; narrow pterygoid flange with facet indicating ectopterygoid with an essentially
anteroposterior orientation; pterygoids separated by large interpterygoid vacuity; presacral
vertebrae procoelous, but notochordal canal closing late in development (at least in middle Jurassic
form); accessory articulations present; axis intercentrum with hypapophysis, but no second
hypapophysis on the rear of the axis centrum; cervical vertebrae short with deep prominent rib
facets and well-developed spines; dorsal centra longer than cervicals; caudal vertebrae (Kirtlington)
ainphicoelous and autotomous.

Parviraptor estesi sp. nov.

Text-figures 1-4, 6a, c, 8a, 9, 10a, 13.

Derivation of name. For the late Dr Richard Estes, in recognition of his work on early lizards and their
relationships.

Holotype. BMNH 48388, a small block bearing an association of skull bones. The specimen was originally part
of the Beckles' collection. The block (Text-figs 1-2) bears parts of the skull of a small lizard including: the left
parietal in ventral view (right in impression); a complete left maxilla in medial aspect; a right pterygoid in
palatal aspect; a left palatine in dorsal aspect; and traces of the left squamosal. Other elements are too
fragmentary for identification. There are isolated teeth and bone fragments near the pterygoid but it is not clear

36

PALAEONTOLOGY, VOLUME 37

text-fig. 3. Parviraptor estesi gen et sp. nov. BMNHR8511; Swanage, Dorset; Purbeck Limestone

Formation; xl.

whether these are of the upper or lower jaw. At one corner of the block is a fragment of what may be a vertebra ;
above it are two pustulate osteoscutes which probably, but not certainly, pertain to the lizard; one is keeled
(Text-fig. 6c). Lying between them is an isolated, rooted, archosaur tooth (see below and Text-fig. 2, arrowed).

Type locality. Durlston Bay, Dorset (precise details unrecorded).

Type horizon : Purbeck Limestone Formation, late Jurassic (Tithonian) or early Cretaceous (Berriasian).

Diagnosis. As for genus, pending detailed comparisons between the specimens from the three
localities.

Referred material. BMNH R851 1, a small slab bearing an association of skull bones (right frontal, left parietal,
left postfrontal or postorbitofrontal, left palatine), ribs and vertebrae (cervical and dorsal) recorded only as
from the Middle Purbeck of Swanage, Dorset (Text-figs 3-4).

EVANS: MESOZOIC LIZARDS FROM ENGLAND

37

text-fig. 4. Parviraptor estesi gen. et sp. nov. BMNHR8551; Swanage, Dorset; Purbeck Limestone
Formation. Key as for Text-figure 2, but in addition. At, atlas arch ; Pf, postfrontal or postorbitofrontal ; x I

Parviraptor cf. P. estesi
Text-figures 6b, 7, 8b-c, 11-12, 14-15.

Material. A series of dissociated elements from the Mammal Bed horizon. Forest Marble (Bathonian),
Kirtlington Cement Works Quarry, Oxfordshire, including: BMNH R 12352, a partial left maxilla; R 12353,
a left parietal; R12354, a partial right dentary; R 12355, the symphysial region of a right dentary; R 12356,
a partial left frontal; R12357, a left frontal of a juvenile; R12358, a left frontal of a juvenile; R12359, a partial
left frontal; R 12360, an axis vertebra; R 12361, a partial dorsal vertebra; R 12362, a partial dorsal vertebra;
R 1 2363, a dorsal vertebra of a juvenile ; R 1 2364, a fragment of a vertebral condyle ; R 1 2365, a dorsal vertebra ;
R12366, an anterior caudal vertebra; R12367, an anterior caudal vertebra; R12368, the anterior part of an
autotomized caudal vertebra; R 12369, the posterior part of an autotomized vertebra. A further collection
of about 100 dissociated jaws, skull bones and vertebrae from the same horizon is held in the Department
of Anatomy and Developmental Biology, University College London.

Comment. There is referable material (uncatalogued) from the late Jurassic (Oxfordian or Kimmeridgian) of
Guimarota, Portugal, in the collections of the Freie Universitat, Berlin.

38

PALAEONTOLOGY, VOLUME 37

Description. The description of Parviraptor is based on the Purbeck specimens in conjunction with
referred material from Kirtlington Quarry, Oxfordshire.

Skull

The skull of Parviraptor is represented by the maxillae, parietals, frontals, a postfrontal or postorbitofrontal,
pterygoids, palatines and dentaries, with a fragment of the squamosal (or possibly supratemporal).
Recognition of these elements permits a partial reconstruction of the skull (Text-fig. 5a-b).

A

B

text-fig. 5. Partial reconstruction of the skull of Parviraptor estesi in A, dorsal, and B, palatal views. Scale bar

represents 10mm.

The maxilla has a long low facial process and shorter premaxillary and suborbital processes (Text-fig. 6a).
The anterodorsal edge of the facial process is smooth and unbroken; it clearly formed an elongated narial
margin in a depressed snout (as confirmed also by specimens from Guimarota). There is no obvious facet for
the prefrontal but a narrow grooved region along the posterodorsal edge must have accommodated this
element. The alveolar border bears around twenty-three tooth positions, most of which are filled. The
implantation is pleurodont but there is a build-up of bone around each tooth base which provides a half socket.
Most of the teeth are broken off near the base, but enough of the tips are preserved to show the distinctive
shape. Each tooth has a round base and a long, sharp, recurved tip - quite different from the broader, almost
triangular teeth of Dorsetisaurus. Only one tooth shows a trace of a replacement pit in the form of a scarred
area on the distolingual surface. This feature, and the occasional presence of a similar excavation in the bone
of attachment suggests an essentially anguimorph mode of replacement. The palatal shelf of the maxilla is
relatively broad, but the lingual edge is smooth and clearly did not meet the vomer (the palaeochoanate
condition). The position of the palatine facet is marked by a weak groove about half-way along the bone-
the edge immediately behind this being broken. However, the length of the facet matches that of the
corresponding process from the palatine, suggesting that the maxilla extended for a short distance into the
margin of the suborbital fenestra. The entry foramen for the maxillary nerve lies close to the facet.

The Kirtlington maxillae are fragmentary and add little to the holotype description. On the medial surface,
however, some specimens show what appears to be a prefrontal facet. This feature is larger than anything
preserved on the holotype maxilla but, by comparison with Portuguese material, the difference is probably due
to damage. R12352 (Text-fig. 6b) is a maxillary fragment which preserves a displaced replacement tooth. This
tooth is similar to those of the holotype and is matched by isolated teeth from the fine sievings.

There is no frontal bone on the holotype block, but BMNH R851 1 from Purbeck preserves a right frontal

EVANS: MESOZOIC LIZARDS FROM ENGLAND

39

text-fig. 6. Parviraptor estesi gen. et sp. nov. a, BMNH 48388, holotype maxilla; lingual view, b, BMNH
R12352; maxillary fragment from Kirtlington. c, BMNH 48388; compound osteoscutes on holotype block.

Scale bars represent 1 mm.

i- i

text-fig. 7. Parviraptor cf. P. estesi. Kirtlington; Forest Marble, (Bathonian). a-d, BMNH R12356; posterior
region of left frontal in a, dorsal, b, ventral, and c, lateral views, d, cross-section, to show development
of subolfactory process, e, BMNH R12357; and f, BMNH R12358; same region in juveniles for comparison
of size, g, BMNH R 12359; lateral view of left frontal to show facets for circumorbital bones. Scale bar

represents 1 mm.

40

PALAEONTOLOGY, VOLUME 37

in lateral view (Text-fig. 9b) showing the facets for the pre- and postfrontal (or postorbitofrontal) and the deep
subolfactory process which stretches toward the midline. Lying against the rear of the subolfactory process is
a second element which appears to be either weakly ossified or simply calcified and is most reasonably
interpreted as the orbitosphenoid. Similar frontals are known from Kirtlington (Text-fig. 7). They vary in size
and are always paired with a simple straight midline suture and little orbital waisting. The dorsal surface is flat,
matt and completely unsculptured. The posterior border is ‘ U ’-shaped, but the frontoparietal joint appears to
have been weak. The subolfactory processes are strongly developed (Text-fig. 7d), but incomplete. The
posterior facet for the postorbitofrontal, or postfrontal, is large, but shallow and mostly ventral. As a result,
the articulation between the postorbitofrontal and frontal is unlikely to have limited mesokinesis.

text-fig. 8. Parviraptor estesi gen. et sp. nov. a, BMNH 48388, holotypc; left parietal and associated
squamosal; ventral view, b-c, BMNH R 12353; left parietal from Kirtlington in B, ventral and c, dorsal views.

Scale bars represent 1 mm.

The parietals (Text-figs 7a, 8a) on the Purbeck blocks are large paired bones with a small parietal foramen.
Each bone is longer than it is broad, and the two parietals together formed a substantial plate (Text-fig. 5a).
The lateral margins are curved and clearly bordered a long upper temporal opening. Running along the
ventrolateral edge is a conspicuous groove, presumably for the taenia marginalis of the chondrocranium. There
are no descending flanges and the jaw adductor muscles do not appear to have been strongly developed;
presumably with the sharp recurved teeth, a quick snapping bite was enough to subdue prey before swallowing.
Posteriorly, each bone tapers laterally into a strong postparietal process and medially into a smaller triangular
process which extended out over the supraoccipital and may have been in contact with it. As preserved, the
postparietal processes bear no obvious facets and the supratemporal or squamosal must have attached further
distally. Anteriorly, the two parietals contribute towards a broad median frontal process while laterally there
is a small, but conspicuous, notch for the posterolateral part of the frontal. In dorsal view (Text-figs 8c, 9a),
the surface is smooth but bears a strong posterior surface for the attachment of cervical muscles.

The largest Kirtlington parietal (R 12353; Text-fig. 8b-c) is complete except for its postparietal process, and
is almost identical to that on BMNH 48388 except that it is proportionally a little shorter. The anterior border
is slightly wider than the posterior one and bears a small shelf facet for the frontal. There is no obvious facet
for the postorbitofrontal, but it probably fitted below the overhanging dorsal rim of the parietal - as it does
in Varanus.

At one end of BMNH R851 1, there is a slender triradiate bone with long anterior and posterior rami and
a short blunt external process (Text-fig. 8d). In its general shape and size, this element could be taken for a
jugal but it lacks a smooth orbital margin and the distribution of facets is incompatible with such an
interpretation. The bone is more plausibly interpreted as a postfrontal, or postorbitofrontal, and can be fitted

EVANS: MESOZOIC LIZARDS FROM ENGLAND

41

text-fig. 9. Parviraptor estesi gen. et sp. nov. Skull bones from BMNH R851 1. a, left parietal with overlying
palatine in outline; b, right frontal in lateral view, c, left palatine in ventral view, d, probable left postfrontal
(or postorbitofrontal) in dorsolateral view. Key: pf, postfrontal facet or (D) region of postfrontal which fits
into this facet; prf, prefrontal facet; Ptp, pterygoid process; Vp, vomerine process; x, possible orbitosphenoid ;

5ii, canal for maxillary nerve. Scale bar represents 1 mm.

against the frontoparietal suture. The faceted anterior ramus thus fits into the facet on the frontal while the
apparently unfaceted posterior ramus presumably had a weaker articulation with the parietal (see above). The
bone is embedded in matrix and its undersurface is not visible. It is unclear how, or even whether, this element
articulated with other circumorbital or temporal bones.

Lying beside the left parietal of the holotype block, partly in impression, is a slender squamosal (Text-fig. 8a).
This lacks a dorsal process and is typically scleroglossan

The left palatine is preserved in dorsal view (Text-fig. 10a) on the holotype and in ventral, palatal, view on
BMNH R8511 (Text-fig. 9c). It is triradiate and roughly as wide as it is long. The slender vomerine process
has a notched anterior border and a long choanal margin. By contrast, the posterior process is short and clearly
had a weak transverse joint with the pterygoid which may have permitted hypokinetic palatal hinging
(Frazzetta 1962). The lateral maxillary process has splayed anterior and posterior limbs. The posterior limb is
slightly raised and bears a groove facet for the anteromedial part of the jugal. It is pierced by a short canal for
the maxillary nerve (Text-fig. 9c). The neck of the palatine, between the maxillary process and palatal plate,
bears a slight rugosity for the prefrontal but this is not prominent (prf. Text-fig. 10a) and the contact may have
been ligamentous. There is no trace of palatine teeth, nor is there a choanal groove or gutter. From the shape
of the palatine, both the choana and the suborbital fenestra were narrow. In its structure and relations, the
palatine of Parviraptor is much like that of a modern Varanus (Text-fig. 10b).

The right pterygoid is preserved in palatal view (Text-fig. 2). It shows a long, tapering quadrate process and
a slender, forked palatal plate. At the medial junction of the two, the recess for the basipterygoid process of
the sphenoid is flanked by a conspicuous knob. Laterally, the pterygoid flange is slightly hook-shaped with a
distinct ectopterygoid facet. The two bones clearly slotted into one another with the ectopterygoid orientated
essentially anteroposteriorly rather than mediolaterally.

The dentary is known only from Kirthngton. The largest dentary fragment is BMNH R 12354
(Text-fig. 1 1 a-b), the midportion of a right bone. The lateral surface is pierced by large, widely spaced sensory
foramina. Medially, the bone has a deep, almost vertical, alveolar shelf with no development of a subdental
ridge. The tooth positions are marked by walls of alveolar bone which clearly built up around the bases of the
teeth, although they do not appear to have held the teeth in place because all specimens are edentulous. The
Meckelian fossa is very shallow and opens ventromedially. As preserved, there is no obvious splenial facet. In

42

PALAEONTOLOGY, VOLUME 37

B

A

text-fig. 10. a, BMNH 48388 ; left palatine of Parviraptor estesi on holotype block, dorsal view, b, left palatine
and associated bones of Varanus sp. Key: ec, ectopterygoid ; J, jugal; Jf, jugal facet; mx, maxilla;
mxp, maxillary process; PI, palatine; prf, prefrontal boss; Ptp, pterygoid process; V, vomer; Vp, vomerine
process; 5ii, course of maxillary nerve. Scale bars represent 1 mm.

text-fig. 11. Parviraptor cf. P. estesi ; Kirtlington; Forest Marble, (Bathonian). A-B, composite right dentary
comprising BMNH R 12355, symphysial region, and BMNH R 12354, midsection, a, lingual, and b, labial
views, c, BMNH R12370 posterior fragment in lingual view to show intramandibular septum (ims). Scale bar

represents 1 mm.

EVANS: MESOZOIC LIZARDS FROM ENGLAND

43

Text-figure 1 1, an isolated symphysial region (R12355) has been included. The symphysis bears a dorsomedial
flange which met the opposite bone in the midline. Ventrolaterally, there is a single large sensory foramen. The
rear end of the tooth row is preserved in BMNH R 12370. Beneath the alveolar border, there is a small
intramandibular septum, separating the entry of the inferior alveolar canal from the Meckelian fossa itself
(Text-fig. 11c).

The postcranial skeleton

Vertebrae. Of the lizard vertebrae at Kirtlington, only one set matches the jaw and skull material in terms of
size and surface texture, and this association is confirmed by BMNH R8511 from Purbeck.

The vertebrae in question are among the most numerous of the lizard vertebrae at Kirtlington and, like the
frontals and parietals, show a considerable size range. They also show what appears to be an interesting
ontogenetic series which correlates with size and illustrates the development of the posterior condyle
(Text-fig. 12d-i). In general, the vertebrae are strongly built, with a large anterior cotyle pierced centrally by a

A

B

C

text-fig. 12. Parviraptor cf. P. estesi. Kirtlington; Forest Marble, Bathonian. a-c, BMNH R 12360; axis
centrum, in ventral, left lateral, and posterior views (Key: h, hypapophysis). d-m, range of dorsal vertebrae
to show size-related stages in the development of the posterior condyle, d-e, BMNH R 12361, dorsal and
posterior views ; f, BMNH R 1 2362, posterior view ; G, R 1 2363 ; posterior view, h-i, BMNH R 1 2364 ; posterior

and dorsal views. Scale bars represent 1 mm.

notochordal canal. The dorsal surface of the centrum bears a conspicuous ridge. The very smallest centra are
free of the neural arches, but small fused vertebrae are fully notochordal (e.g. Text-fig 12g). In slightly larger
specimens, a lip of bone appears on the posteroventral edge of the centrum (e.g. Text-fig. 12d-e). This increases
in size, spreading upward to form a grooved process. When fully developed, the sides of the grooves have met
in the midline leaving a condyle with a small hole or pit to mark the original position of the notochord
(Text-fig. 12h-i). This is the development process followed in the development of lizard procoely (Winchester
and Bellairs 1977), but it is unusual to find all stages within a series of posthatchling vertebrae (except in
procoelous geckos and xantusiids). Even when the condyle is fully formed, the notochordal canal remains open
within the remainder of the vertebra.

44

PALAEONTOLOGY, VOLUME 37

This apparent ontogenetic series is interesting. Each of the other lizards at Kirtlington shows a notable
uniformity of size in its individual elements, as would be expected in a group normally characterized by
determinate growth. This is clearly not true of Parviraptor where the vertebrae, as well as the frontals and
parietals, despite a constant morphology, show a range of sizes which may indicate a relatively long period of
growth before reaching adult size. It raises the question as to whether any of the Kirtlington specimens are
actually adult. The largest Kirtlington parietals are only about half the size of those on the holotype block,
while some of the Guimarota bones are two or three times the size of comparable Purbeck elements.
Furthermore, the vertebrae on BMNH R8511, although of similar size to those from Kirtlington, have fully
developed condyles with no trace of the notochordal canal. Whether these variants represent several different-
sized species or a series of developmental stages within a single species is impossible to determine at the present
time.

On the basis of specimens from both Kirtlington and Purbeck, representative elements from most
regions of the vertebral column have been identified. An isolated atlas arch is preserved on BMNH R8511
(Text-fig. 14a), but the only known axis vertebrae are those from Kirtlington. The axis (Text-fig. 12a-c) bears
a small odontoid process and there is a prominent hypapophysis on the intercentrum. The posterior end of the
centrum is notochordal, but there is some development of the ventral lip. There is, however, no trace of a
second, posteriorly placed, hypapophysis.

text-fig. 13. Parviraptor estesi. BMNH R851 1 ; Swanage, Dorset; vertebrae, a, right atlas arch in lateral view.
B, cervical vertebra in left ventrolateral view, c, fragmentary cervical vertebra in end view to show deep rib
facets and neural spine, d, cervical vertebra in dorsal view, e, dorsal vertebra in right lateral view, f, dorsal
vertebra, dorsal view. Key: r, rib process. Scale bar represents 1 mm.

Cervical vertebrae are preserved on BMNH R8511. They are shorter than the dorsals and bear deep,
prominent rib facets (Text-fig. 13b-d) and well-developed neural spines. Unfortunately, none is sufficiently well
preserved in ventral view to determine the presence or position of cervical hypapophyses. Dorsal vertebrae are
known from both Kirtlington and Purbeck (Text-figs 13e-f, 14). The centrum is broad, almost rectangular
(Text-fig. 14b), and lacks subcentral foramina. The neural arch is wide and bears large, flared anterior and
posterior zygapophyses with a rudimentary zygosphene/zygantrum system (Text-fig. 14a,d), but there seems
to be little development of the neural spine.

Caudal vertebrae are known only from Kirtlington. Anterior caudal vertebrae bear transverse processes and
typical neural spines (Text-fig. 15a-f). The midventral groove becomes rugose and pitted as the region of
autotomy is approached. The presence of several hemivertebrae showing the characteristic midline transverse

EVANS: MESOZOIC LIZARDS FROM ENGLAND

45

text-fig. 14. Parviraptor cf. P. estesi. BMNH R12365; Kirtlington; Forest Marble, (Bathonian). Dorsal
vertebra, in dorsal, ventral, posterior, anterior, and left lateral views. Dotted lines in E show overlap between

successive vertebrae. Scale bar represents 1 mm.

text-fig. 15. Parviraptor cf. P. estesi. Kirtlington; Forest Marble, (Bathonian). a-e, BMNH R 12366; anterior
caudal vertebra; in left lateral, dorsal, posterior, anterior, and ventral views, f, BMNH R 12367; anterior
caudal in ventral view, showing groove for caudal blood vessels, g-h, fragments of autotomous vertebra
G, BMNH R12368; short portion anterior to fracture plane; H, BMNH R12369; longer portion posterior
to fracture plane (cranial end uppermost in each case). Scale bars represent 1 mm.

46

PALAEONTOLOGY, VOLUME 37

ridge confirms that the tail contained functional autotomy planes. There were apparently no transverse
processes on autotomous vertebrae but the fracture plane divided the vertebra into a small anterior component
and a longer posterior one (Text-fig. 1 5g-h). The absence of any complete middle or posterior caudals suggests
that most elements of the tail retained autotomy septa. The caudal vertebrae share the thick central lips of the
presacrals, but remain notochordal.

Ribs. A set of strong single-headed ribs is preserved on BMNH R8511 (Text-figs 3-4).

Discussion. The fully pleurodont dentition, the simple arched squamosal and the appearance of a
mesokinetic hinge line in the palate and skull roof suggest that Parviraptor was a crown-group
lizard. This conclusion is further supported by the relatively short jaws (primitive relatives typically
have long narrow jaws; Evans 1980, 1991) and procoelous vertebrae.

Determining the position of Parviraptor within lizards is more complicated, since (following Estes
1983; Estes et at. 1988) it appears to show a combination of gekkotan and anguimorph character
states. Kluge’s detailed analyses of modern gekkotans (e.g. 1967, 1987) have suggested that paired
parietals and the posthatchling retention of a notochordal canal are synapomorphies of Gekkota,
resulting from paedomorphosis. Consequently, fossil lizards exhibiting one or other of these
character states have frequently been classified as gekkotans (Estes 1983). However, some of the
same character states (paired parietals, free trunk intercentra, centrum development) are also
present in living xantusiids, once classified with Gekkota but now generally regarded as
scincomorphs (Estes et al. 1988), and some specimens of Varanus retain a midline suture in the
anterior part of the parietal (personal observations). Clearly then, the same states have arisen more
than once within crown-group lizards. The retention of paired parietals in Parviraptor , and the late
formation of the vertebral condyle, are just as likely to be the result of its extended growth period
as an indication of gekkotan relationship.

Taking this into account, the majority of the derived characters displayed by Parviraptor suggest
anguimorph, rather than gekkotan, affinity (Borsuk-Bialynicka 1984; Pregill et al. 1986; Estes et al.
1988):

(a) intramandibular septum (although a small septum can be present in iguanians and lacertoids)

(b) shallow pleurodont implantation with bone of attachment; replacement pits lost or reduced -
tooth replacement may be distolingual

(c) deep lingual shelf but no subdental ridge

(d) Meckel’s canal opens anteroventrally

(e) probably compound osteoscutes (although these are found in a few other groups).

There is, however, a problem. In all living anguimorphs, cervical hypapophyses attach to the rear
of the preceding centrum. They are fused to the centrum in anguids and sutured to it in varanoids.
The preservation of this region of the column in Parviraptor is admittedly poor, but it is disturbing
that no trace of hypapophyses, nor of a sutural surface for a hypapophysis, is visible behind the
level of the atlas intercentrum.

Within Anguimorpha, Parviraptor is strikingly different from Dorsetisaurus in its sharp recurved
teeth, the low maxilla, the presence of an intramandibular septum, the absence of dermal sculpture
on the roofing bones, the structure of the vertebrae, the shape of the parietals and the apparently
extended growth period.

Comparison with both living and extinct anguimorph genera, using principally the character sets
of Borsuk-Bialynicka (1984), Pregill et al. (1986) and Estes et al. (1988), produces conflicting
hypotheses. In a number of features, Parviraptor resembles platynotan lizards:

(a) the absence of osteodermal incrustation on the skull bones (osteoderms and dermal ornament
are usually free of the skull bones in platynotans. Heloderma may be specialized in this respect;
it has a tesselate arrangement of cranial osteoderms which become attached to the underlying
bones)

text-fig. 16. Cladogram showing hypothetical relationships for Parviraptor based on the work of Estes (1983),
Borsuk-Bialynicka (1984) and Pregill et al. (1986). Key: Node 1, Anguimorpha (intramandibular septum;
replacement teeth usually posterolingual; loss/reduction of subdental ridge; Meckel's groove opens
anteroventrally ; dorsal and cephalic osteoderms); Node 2, Platynota (sertsu Pregill et al. 1986) (well-developed
subolfactory processes on frontals; teeth unicuspid, recurved, trenchant; reduction in maxillary/nasal contact;
vomers narrow, more than twice palatine length; palatine as wide as it is long; osteoderms free of skull bones);
Node 3, unnamed taxon (parietals fused; adductor muscles extend onto dorsal surface of parietal;
splenial/dentary overlap reduced; vertebral condyle oblique [?R in aquatic genera] and waisted; postaxial
cervical centra with fused hypapophyses bearing separate epiphyses); Node 4, unnamed taxon - aquatic genera
denote aigialosaurs and mosasaurs (retracted nares; frontal trapezoid; maxilla does not underlie orbit; facial
process of maxilla set posteriorly; plicidentine present [varies in 'necrosaurs']; reduced overlap of dentary and
postdentary bones; separate epiphyses on cervical hypapophyses [varies in ‘necrosaurs’]; loss of caudal
autotomy); Node 5, Varanoidea {sensu Pregill et al. 1986, i.e. Varanidae, Lanthanotidae, Helodermatidae)
(subolfactory processes of frontals usually meet in the midline; fewer than 13 maxillary teeth; supratemporal

extends anterior to level of parietal notch).

(b) a palatine that is as wide as it is long, has no choanal groove, lacks denticles, and has a weak
prefrontal boss suggestive of a ligamentous attachment between the palatine and prefrontal

(c) the long slender vomers (as deduced from the reconstruction - but since the length of the
maxilla, the position of its facet with the palatine, and the shape of the palatine are all known,
this is not speculative)

(d) the wide interpterygoid vacuity

(e) the sharp, unicuspid, strongly recurved teeth

(f) the depressed form of the maxilla with its elongated narial margin

(g) the strongly developed subolfactory processes.

There are, however, character conflicts and Parviraptor lacks a number of character states found in

living varanoids (Pregill et al. 1986; Estes et al. 1988) notably;

(a) the presence of plicidentine (infolding at the tooth base)

(b) the invasion of the dorsal surface of the parietal adductor musculature

(c) the presence of sutural surfaces at the rear of each cervical vertebra for the attachment of
hypapophyses

(d) the anterior elongation of the supratemporal to reach the level of the parietal notch (the point
at which the postparietal processes leave the skull table)

48

PALAEONTOLOGY, VOLUME 37

(e) facial process of maxilla set towards rear of bone, premaxillary process horizontally expanded
(0 small number (fewer than 13) of well-spaced maxillary teeth

(g) the obliquity of the vertebral condyles

(h) the waisting of the vertebral condyles

(i) the loss of autotomy septa.

These character states have arisen within Platynota ( sensit Pregill et al. 1986) and, in most cases,
their absence in a very early representative would not present great difficulties. Text-figure 16
illustrates currently agreed relationships amongst the Platynota (Borsuk-Bialynicka 1984; Pregill
et al. 1986) based on derived character states. Parviraptor emerges as having a basal position
amongst the poorly resolved ‘ necrosaurs stem platynotans known from the late Cretaceous to the
Oligocene (Estes 1983; Pregill et al. 1986). As currently constituted, necrosaurs form a grade rather
than a clearly diagnosed monophyletic group, and it is likely that some genera lie closer to the crown
than others. Thus the presence of plicidentine in some genera but not others (Estes 1983) is as likely
to be due to its primitive absence in some genera as to secondary loss. Some necrosaurs (e.g.
Necrosaurus from the Palaeocene-Oligocene of Europe; Estes 1983) show a build up of attachment
bone around the tooth bases similar to that of Parviraptor.

If Parviraptor is genuinely platynotan, its presence in the middle Jurassic is not surprising (and
was predicted by Borsuk-Bialynicka 1984, p. 67) since the appearance of the possible anguioid
Dorsetisaurus in the late Jurassic (Portugal, North America) indicates that the primary anguimorph
dichotomy had already taken place.

THE ISOLATED ARCHOSAUR TOOTH

The small isolated archosaur tooth on the holotype block (Text-fig. 2, arrowed) has a compressed
conical crown and a root which is slightly wider than the crown. Unlike the small crocodilian teeth
found at Purbeck, however, the crown does not bear striae. In its general shape, the tooth resembles
those of the genus Lisboasaunts from the late Jurassic of Portugal (Milner and Evans 1991, fig. 6).
Seiffert (1973) described Lisboasaurus as an anguimorph, but Estes (1983) noted the resemblance
between the type maxilla and that of a small therapod. Milner and Evans (1991) reached the same
conclusion, suggesting that the genus might be related to troodonts or, possibly, birds. If the
Purbeck tooth is attributable to Lisboasaurus , it would provide another link between the Jurassic
assemblages of Britain and Portugal.

Acknowledgements. I thank Dr David Unwin who drew my attention to the presence of BMNH R8551 in the
pterosaur collections of The Natural History Museum. I thank Sandra Chapman and Dr Angela Milner at the
Natural History Museum, London, for arranging the loan of the Purbeck specimens; Professor Bernard Krebs,
Freie Universitat Berlin, for access to Guimarota material; and Dr Olivier Rieppel, Chicago, for helpful
criticism of the text. This work was partly funded by a grant from the University of London Central Research
Fund. The Kirtlington material was collected by a team from University College under the direction of
Professor Kenneth Kermack and 1 am grateful to him for his continued interest. The Photography Section,
Department of Palaeontology, The Natural History Museum, London, provided the photographs which form
Text-figures 1 and 3.

REFERENCES

allen, p. and wimbledon, w. a. 1991. Correlation of NW European Purbeck - Wealden (non-marine Lower
Cretaceous) as seen from the English type areas. Cretaceous Research , 12, 511-526.
borsuk-bialynicka, m. 1984. Anguimorphans and related lizards from the late Cretaceous of the Gobi Desert,
Mongolia. Palaeontologia Polonica. 46, 5-105.

carroll, r. l. and debraga, m. 1992. Aigialosaurs : mid-Cretaceous varanoid lizards. Journal of Vertebrate
Paleontology , 12, 66-86.

estes, R. 1983. Sauria Terrestria , Amphisbaenia. In wellnhofer, p. ( ed . ) . Handbuch der Paldoherpetologie , 10A.
Gustav Fischer Verlag, Stuttgart, 249 pp.

EVANS: MESOZOIC LIZARDS FROM ENGLAND

49

— de queiroz, k. and gauthier, j. 1988. Phylogenetic relationships within Squamata. 1 19-282. In estes, r.
and pregill, G. (eds). Phylogenetic relationships of the lizard families. Stanford University Press, Stanford,
California, 631 pp.

evans, s. e. 1980. The skull of a new eosuchian reptile from the Lower Jurassic of South Wales. Zoological
Journal of the Linnean Society, 70, 203-264.

— 1991. A new lizard-like reptile (Reptilia: Lepidosauromorpha) from the Middle Jurassic of Britain.
Zoological Journal of the Linnean Society , 103, 391 — 41 2.

— (in press). Jurassic lizard assemblages. Second Georges Cuvier Symposium. Revue de Paleobiologie. 7.
frazzetta, t. h. 1962. A functional consideration of cranial kinesis in lizards. Journal of Morphology, 111,

287-320.

furbringer, m. 1900. Zur vergleichenden Anatomie des Brustschulterapparates und der Schultermuskeln. IV
Teil. Gegenbaurs Morphologische Jahrbuch , 34, 215-718.
hoffstetter, r. 1964. Revue des recentes acquisitions concernant Lhistoire et la systematique des Squamates.
Colloques Internationaux du Centre National de la Recherche Scientifique, 104, 243-279.

— 1967. Coup d’oeil sur les Sauriens ( = lacertiliens) des couches de Purbeck (Jurassique superieur
d’Angleterre). Colloques Internationaux du Centre National de la Recherche Scientifique, 163, 349-371.

kluge, a. G. 1967. Higher taxonomic categories of gekkonid lizards and their evolution. Bulletin of the
American Museum of Natural History, 135, 1—59.

— 1987. Cladistic relationships in the Gekkonoidea (Squamata, Sauria). Miscellaneous Publications,
Museum of Zoology, University of Michigan , 173, 1-54.

kuhn, o. 1958. Ein neuer Lacertilier aus dem frankischen Lithographieschiefer. Neues Jahrbuch fur Geologie
und Paldontologie, Monatshefte, 1958, 380-382.

milner, a. r. and evans, s. e. 1991 . The Upper Jurassic diapsid Lisboasaurus estesi - a maniraptoran theropod.
Palaeontology, 34, 503-513.

OPPEL, M. 1811. Die Ordnungen, Familien und Gattungen der Reptilien , als Prodr om einer Naturgeschichte
derselben. Joseph Lindauer, Munich, 87 pp.

pregill, g., gauthier, j. and greene, h. 1986. The evolution of helodermatid squamates, with description of
a new taxon and an overview of Varanoidea. Transaction of the San Diego Society of Natural History, 21,

1 67-202.

prothero, d. r. and estes, R. 1980. Late Jurassic lizards from Como Bluff, Wyoming, and their
palaeobiogeographic significance. Nature, 286, 484-486.
seiffert, j. 1973. Upper Jurassic lizards from Central Portugal. Memoria, Servigos Geoldgicos de Portugal , 22,
1-85.

winchester, l. and bellairs, a. d’a. 1977. Aspects of vertebral development in lizards and snakes. Journal of
Zoology, 181, 495-525.

SUSAN E. EVANS

Department of Anatomy and Developmental Biology
University College London

Typescript received 16 October 1992 Gower Street

Revised typescript received 19 February 1993 London WC1E 6BT, UK

THE LOWER PERMIAN SYNAPSID GLAUCOSAU RUS

FROM TEXAS

by S. P. MODESTO

Abstract. The early synapsid Glaucosaurus megalops , from the Lower Permian of north-central Texas, is re-
examined. Despite being represented by a single, partial skull of uncertain ontogenetic age, the presence of six
autapomorphies indicates that Glaucosaurus is clearly a distinct synapsid form. Phylogenetic analysis of the
early Synapsida indicates that Glaucosaurus is the probable sister taxon of Edaphosaurus within
Edaphosauridae. The clade of Glaucosaurus and Edaphosaurus is distinguished from its sister taxon
Ianthasaurus by five synapomorphies : (1) premaxillary teeth equal to maxillary teeth in size, (2) caniniform
region absent, (3) caniniform tooth absent, (4) prefrontal ventral process expanded transversely, and (5)
transverse flange of pterygoid absent.

In 1915 S. W. Williston erected the taxon Glaucosaurus megalops for a small partial skull recovered
from the Mitchell Creek locality (Waggoner Ranch Formation, Wichita Group, Lower Permian)
of Baylor County, Texas. He remarked that the skull of Glaucosaurus resembled that of the large
edaphosaurid Edaphosaurus. However, the original preparation was poor and resulted in the
elimination of many sutures. The partial preservation of the type and the unfortunate loss of
discernible sutures has made Glaucosaurus megalops one of the most problematical of Permo-
Carboniferous synapsid taxa.

Despite the fragmentary nature of the holotype, Broom (1932) re-examined it and allied
Glaucosaurus with another early synapsid from Mitchell Creek, Mycterosaurus longiceps, and a
South African form, Anningia megalops ; the order Anningiamorpha was erected to encompass these
three taxa (Broom 1932). Mycterosaurus has since been recognized as a varanopseid eupelycosaur
(Berman and Reisz 1982), whereas Anningia and the order bearing its name were declared nomina
vana (Reisz and Dilkes 1992).

In their monumental review of Permo-Carboniferous synapsids, Romer and Price (1940) noted
that Glaucosaurus had little in common with any early synapsid other than Edaphosaurus , and
identified Glaucosaurus as a primitive member of their (polyphyletic) synapsid suborder
Edaphosauria, suggesting further a close relationship with Casea. This hypothesis was prompted by
their assumption that both taxa had a single coronoid. However, Casea possesses two coronoids
(Sigogneau-Russell and Russell 1974), and the fragmentary lower jaws of Glaucosaurus were only
partially prepared; the sutural patterns of the mandible of Glaucosaurus could not have been
determined faithfully in medial view.

More recently, Reisz ( 1986) concluded that Glaucosaurus possessed none of the apomorphies that
defined the six major groups of Permo-Carboniferous synapsids, and regarded Glaucosaurus as a
synapsid of uncertain phylogenetic affinities. He reiterated the remarks of Williston (1915), and
Romer and Price (1940) that the general shape of the skull of Glaucosaurus resembled that of
Edaphosaurus , but cautioned that the lack of discernible sutures would make its assignment to any
of the established synapsid families quite tentative.

However, several studies in the past few years have greatly expanded our knowledge of the
anatomy and interrelationships of eupelycosaurian synapsids (Reisz et al. 1992), including
edaphosaurids (Reisz and Berman 1986; Modesto 1991; Modesto and Reisz 1992). Work on the
oldest known edaphosaurid Ianthasaurus (Modesto and Reisz 1990) indicates that edaphosaurids

| Palaeontology, Vol. 37, Part I, 1994, pp. 51— 60|

© The Palaeontological Association

52

PALAEONTOLOGY, VOLUME 37

are united only by the distinctive morphology of their presacral neural spines. Hence, although
Glaucosaurus is clearly not referable to most Permo-Carboniferous synapsid families, there is no
cranial evidence excluding it from the Edaphosauridae. Since Modesto and Reisz (1992) asserted
that the adaptation to terrestrial vertebrate herbivory arose within the edaphosaurid clade, the
possibility that Glaucosaurus is a close relative of Edaphosaurus within Edaphosauridae is intriguing,
and may provide evidence of the development of herbivory in the latter genus.

The holotype of Glaucosaurus megalops is therefore re-examined in order to determine if it is
indeed a relative of Edaphosaurus , or if it represents a distinct eupelycosaurian form, with a
phylogenetic position elsewhere on the synapsid tree.

MATERIALS

The holotype is the only specimen assigned to the taxon Glaucosaurus megalops. The teeth,
antorbital margin, external nares, and inner aspect of the mandibles were prepared further with
grinder and pin-vice. Preparation of the ventral aspect of the palate was hindered by the in situ
preservation of the mandibles, and also by damage resulting from a drillhole, which had been bored
into the skull from below (for mounting purposes).

For the phylogenetic review, specimens assigned to the synapsid taxa Eothyris parkeyi (MCZ
1161), Edaphosaurus novomexicanus (FMNH UC 674), and Edaphosaurus boanerges (MCZ 1762)
were examined.

Institutional abbreviations are: FMNH UC, Field Museum of Natural History, University of
Chicago; MCZ, Museum of Comparative Zoology, Harvard University.

SYSTEMATIC PALAEONTOLOGY

synapsida Osborn, 1903
eupely cosauri a Kemp, 1982
Family edaphosauridae Cope, 1882
Genus glaucosaurus Williston, 1915

Type species. Glaucosaurus megalops Williston, 1915.

Diagnosis. Short-snouted edaphosaurid eupelycosaur distinguished from all other synapsids by the
presence of a prearticular that extends anteriorly as far as the jaw symphysis. Glaucosaurus is
distinguished from other edaphosaurids by the following autapomorphies : septomaxilla exposed
facially; three premaxillary teeth; maxilla long, extends posteriorly to posterior orbital margin;
marginal teeth compressed laterally; subtemporal bar deeper dorso-ventrally than suborbital bar.

Glaucosaurus megalops Williston, 1915
Text-figures 1-3

1915 Glaucosaurus megalops Williston, p. 576, fig. 1.

1932 Glaucosaurus megalops ; Broom, p. 15, fig. 2e-g.

1940 Glaucosaurus megalops ; Romer, and Price p. 507, pi. 20.

Diagnosis. As for genus, this being the only recognized species.

Locality and Horizon. Mitchell Creek, near Maybelle, Baylor County, Texas; Waggoner Ranch Formation
(formerly Clyde Formation; see Hentz (1988) for stratigraphical review of north-central Texas), Wichita
Group, Lower Permian.

Holotype. FMNH UC 691, the only known specimen.

Description. The skull (Text-fig. 1) is compressed obliquely and slighlly transversely. Much of the skull roof,
posterior cheek, occiput, and posterior ends of the mandibles are missing, and the braincase is absent. Most

MODESTO: LOWER PERMIAN SYNAPSID

53

text-fig. 1. Glaucosaurus megalops Williston, 1915. FMNH UC 691, holotype. a, left lateral view of skull;
b, posterolateral view of left antorbital region showing antorbital buttressing; c, right lateral view of skull;
d, dorsal view of skull; E, ventromedial view of left mandible. Most of the sutures on the skull table
were obliterated by previous preparators; for the purpose of clarity, their probable positions are not indicated.

Scale bar represents 10 mm.

elements forming the dorsal margin of the orbits are preserved partly as impression, albeit poorly, but still give
a fair indication of the size of the orbit. The surfaces of many elements, particularly the nasals, prefrontals, and
lacrimals, are damaged by varying degrees of overpreparation. The palate is visible largely in dorsal view.

54

PALAEONTOLOGY, VOLUME 37

Although the relatively large orbits suggest immaturity, the ontogenetic age of the skull is uncertain as no
endochondral elements are preserved.

The redescription permits a revised reconstruction of the skull in lateral view (Text-fig. 2). The unpreserved
portions are restored using edaphosaurid sutural patterns. No reconstruction of the skull is offered in dorsal
view, as many of the elements of the skull roof are either absent or damaged, and few of their sutures can be
discerned faithfully.

text-fig. 2. Glaucosaurus megalops. Restoration of skull roof in left lateral view. Scale bar represents 10 mm.

The premaxilla (Text-fig. 1a, c) resembles those of other eupelycosaurs, but the anterior ascending process
is damaged and it is unclear how far posterodorsally it extended. Three teeth are present (contra Williston
1915). Judging from their basal diameters, these teeth were similar to the maxillary teeth in size.

The maxilla is a long, low element that extends posteriorly to the level of the postorbital bar (Text-fig. 1a).
Despite crushing, its ventral margin appears to have been slightly convex in lateral aspect. There are fourteen
teeth present in the complete left maxilla, with a position for one more, indicating that approximately eighteen
teeth were present in each jaw. Although the labial surfaces have been planed away from most of the maxillary
teeth, what remains of their profile, together with their subequal basal diameters, indicates that neither a
caniniform tooth nor a caniniform region was present. The anteriormost teeth on the right maxilla are laterally
compressed, slightly recurved, and sharply tipped. In contrast, the posteriormost tooth of the left maxilla is
peg-shaped, and lacks the curvature and transverse flattening of the anterior teeth, which suggest that the
posteriormost maxillary teeth were slightly more robust than the anterior teeth. There is no septomaxillary
foramen in the expected position immediately posterior to the base of the septomaxilla. There is, however, an
elongate perforation in the right maxilla immediately posterior to the septomaxilla, but it is unclear whether
this opening represents either surface damage or a foramen the size of which has been exaggerated by
overpreparation.

As in other early synapsids, the septomaxilla is a robust, vertical bone with a medial process that occupies
the posterior half of the external naris (Text-fig. 1a). The septomaxilla is exposed facially, as its lateral surface
is clearly confluent with that of the nasal, despite minor surface damage to the latter. Broom (1932) also
interpreted a facial exposure for the septomaxilla, but restored it incorrectly as an elongate triangle of bone
interposed between the lacrimal and the maxilla.

The triradiate jugal (Text-fig. 1a) resembles those of Ianthasaurus and Haptodus in general aspects, except
that its subtemporal process is deeper dorso-ventrally than the suborbital process, and its ventral exposure is
restricted between the maxilla and the quadratojugal. The jugal is slightly arched posteroventrally, indicating
that the posterior cheek was emarginated as in edaphosaurids, sphenacodonts (sensu Reisz et a/. 1992), and
most ophiacodontids. The lateral surface is lightly sculpted with small, circular dimples. Along the ventral edge
of the subtemporal bar of the jugal there appears to be a small fragment of bone, which previous investigators
have identified as the anterior tip of the quadratojugal. If this is indeed the quadratojugal and not simply a
product of crushing, then the quadratojugal would have formed the ventral edge of the subtemporal bar.
A partial postorbital is preserved in association with the postorbital ramus of the jugal ; it appears to be
displaced slightly posteriorly.

MODESTO: LOWER PERMIAN SYNAPSID

55

Little can be said of the lacrimal except that it probably extended between the septomaxilla and the orbit,
and forms the lateral portion of the antorbital buttress. Very little can be made of the surfaces and the margins
of the nasals and the fragmentary frontals. The sculpturing on the skull table commented on by Lewis and
Vaughn (1965) has been exaggerated by overpreparation.

The external surfaces of both prefrontals have been damaged by poor preservation and crushing, and their
sutures with the nasals and lacrimal cannot be determined with certainty. The ventral process of the prefrontal,
seen in posterior aspect and in frontal section (Text-fig. Ib, d), is greatly expanded transversely and appears
to form most of the antorbital buttress. Overpreparation has removed the lacrimal-prefrontal suture, but the
general shape of the ventral process of the prefrontal suggests that it contacted the palatine, as in Edaphosaurus .

An imperfectly preserved portion of bone is located posteriorly on the skull table where one would expect
the parietal and the postparietal to lie (Text-fig. 1c-d). If identified correctly and not displaced too far
anteroventrally by crushing, then this fragment suggests that the occiput sloped anterodorsally, as in other
early synapsids.

The dorsal surface of the right half of the palate has been damaged by overpreparation and it is impossible
to describe surface detail or sutural patterns except the probable suture with the maxilla. However, the dorsal
surface of the left half of the palate is better preserved (Text-fig. 1b) and reveals that the palatine has a
prominent dorsal process. Unfortunately, damage by previous preparators has obscured the sutures with the
lacrimal and the prefrontal, and the suture between the palatine and the ectopterygoid is obscured by crushing.
The dorsal process of the pterygoid (Text-fig. 1a) is dorso-ventrally low, as in Edaphosaurus. The palatal ramus
of the pterygoid is arched slightly ventrally in parasagittal section, and there is no suggestion of a transverse
flange. The palatal process does not appear to bear any teeth posteriorly, and only slight wear is present on
its posterolateral edge (Text-fig. lc).

Anteriorly, the dentary (Text-fig. lc) is relatively deep as in caseasaurs and sphenacodonts, and not
acuminate as reported by Williston (1915) and Romer and Price (1940). Only the last two dentary teeth are
visible (Text-fig. 1a); they differ from the last maxillary tooth only by their smaller size.

The splenial (Text-fig. lc, e) is exposed in lateral aspect only at the jaw symphysis and where it underlies the
anterior tip of the angular. Medially, the splenial extends about halfway up the medial aspect of the mandible.
The lingual surface of the left splenial has been subject to some crushing, and it is uncertain whether an
inframeckelian foramen was present ; the partial right splenial is uncrushed and shows no indication of such
an opening anteriorly.

The prearticular is exceedingly long, extending posteriorly from the jaw symphysis (Text-fig. 1 e). It increases
slightly in dorsoventral height as one progresses posteriorly, at least as far as its preserved posterior end at the
level of the last dentary tooth. Dorsal to the prearticular is a very slender splint of poorly preserved bone, which
probably represents the anterior coronoid and the partial posterior coronoid. None of these elements possesses
teeth.

PHYLOGENETIC RELATIONSHIPS

Re-examination of the holotype of Glaucosaurus megalops has uncovered several features that were
overlooked by previous workers. Glaucosaurus possesses a suite of autapomorphies (see diagnosis)
which indicates that this form cannot be recognized as a juvenile of any other synapsid taxon.
Glaucosaurus also features two apomorphies (tall antorbital region, cheek concave posteriorly) that
suggest a phylogenetic position within Eupelycosauria. In order to determine the most probable
position of Glaucosaurus megalops within Synapsida, the interrelationships of the early synapsid
families must be reexamined.

The following synapsid taxa, including Glaucosaurus , form the ingroup. Among caseasaurs, the
caseid Cotylorhynchus (Romer and Price 1940; Stovall et al. 1966) is used, with additional
information taken from Casea (Sigogneau-Russell and Russell 1974). Eothyris represents the family
Eothyrididae; information on this form is taken from the holotype. The families Varanopseidae and
Ophiacodontidae are represented by Mycterosaurus (Berman and Reisz 1982) and an undescribed
ophiacodontid from the Upper Pennsylvanian of Kansas (Wilson 1989), respectively. Two
edaphosaurid genera are used : Ianthasaurus (Modesto and Reisz 1990) and Edaphosaurus (Modesto
1991 ; Modesto and Reisz 1992). Haptodus (Laurin 1993) serves as a representative sphenacodont.

Two outgroups are used. The seymouriamorph Seymouria serves as the first outgroup (White
1939; Berman et al. 1987). The captorhinid reptile Captorhinus represents the second outgroup. The
anatomy of Captorhinus (distinguished from ‘ Eocaptorhinus' only by its multiple-tooth rowed

56

PALAEONTOLOGY, VOLUME 37

dentition) is well known from several recent papers (Holmes 1977; Heaton 1979; Dilkes and Reisz
1986).

Thirty-four characters were used in the analysis. The majority of these were taken from the
literature (Brinkman and Eberth 1983 ; Reisz 1986; Gauthier et al. 1988), but a few are original. The
analysis was run on a Macintosh Ilsi computer using the branch-and-bound algorithm of PAUP
3.0, which is guaranteed to find the most parsimonious trees (Swofford 1989). All characters were
optimized using DELTRAN, and run unordered.

Only one most parsimonious tree was found (Text-fig. 3). It requires sixty-one steps and has a
consistency index of 0 60 excluding uninformative characters. The cladistic analysis supports the
hypothesis of Reisz (1986) concerning the interrelationships of the six early synapsid families;
characters diagnosing all nodes except node G have been discussed elsewhere (Reisz 1986; Modesto
and Reisz 1992) and are not discussed here.

Interestingly, the tree indicates that Glaucosaurus is the sister taxon of Edaphosaurus within
Edaphosauridae. This relationship is relatively robust, as three more steps would be necessary
to remove Glaucosaurus from Edaphosauridae. This sister-group relationship is supported by
the presence of five synapomorphies, which are described below. The number of the character
is enclosed in parentheses; a character preceded by a minus sign indicates that it is a reversal,
and ambiguous characters are denoted by an asterisk :

Premaxillary teeth equal to maxillary teeth in size (-2*). On the basis of basal cross-sectional
diameter, the first premaxillary tooth is roughly equal to that of an average maxillary tooth in both
Glaucosaurus and Edaphosaurus. However, because the premaxilla is known in Ianthasaurus , this
character may apply at a more inclusive node. The presence of premaxillary teeth larger than
maxillary teeth (except caniniforms) is primitive for synapsids.

Caniniform region absent (3). Neither Glaucosaurus nor Edaphosaurus features a caniniform
region. The presence of a caniniform region, found in all other taxa examined here except caseids,
represents the primitive condition.

Caniniform tooth absent (4). There is no caniniform tooth in either Glaucosaurus or Edaphosaurus .
The presence of a caniniform, defined here as a tooth whose basal diameter is notably greater than
that of other maxillary teeth, represents the primitive condition.

Prefrontal ventral process expanded transversely (10). In Glaucosaurus and Edaphosaurus the
ventral process of the prefrontal is greatly expanded transversely, and forms most of the antorbital
buttress that is present in both taxa. The primitive condition of a prefrontal ventral process that
is not expanded transversely is found in most other synapsids, including Ianthasaurus. Eothyris
is interpreted here to have evolved the derived condition convergently.

Pterygoid transverse flange absent (22*). The pterygoid of Glaucosaurus and Edaphosaurus lacks
a transverse flange. Because the palate is unknown for Ianthasaurus , this character may diagnose
a more inclusive node. The presence of a transverse flange on the palatal ramus of the pterygoid
is primitive for all other taxa examined here.

The absence of both the back of the skull and attributable postcrania has obscured past attempts
to place Glaucosaurus among early synapsids. Although it has been suggested (Romer and Price
1940; Lewis and Vaughn 1965) that some of the postcrania from the Mitchell Creek locality,
attributed originally to Mycterosaurus and clearly not edaphosaurid in nature, might belong to
Glaucosaurus , this material was referred confidently to Mycterosaurus and to an unidentified
temnospondyl amphibian by Berman and Reisz ( 1982). Traditionally, the absence of postcrania has
excluded Glaucosaurus from the Edaphosauridae, as edaphosaurids are united only by their
distinctive presacral neural spine morphology (Modesto and Reisz 1990). However, the fact that

MODESTO: LOWER PERMIAN SYNAPSID

57

edaphosaurids are united only by a few, albeit highly distinctive, postcranial characters is due partly
to our inadequate knowledge of the oldest known member of the family, Ianthasaurus hardestii. The
braincase, the palate, and most of the appendicular skeleton of this edaphosaurid are unknown
(Reisz and Berman 1986; Modesto and Reisz 1990). Hence, the description of new Ianthasaurus
material (and, similarly, the discovery of Glaucosaurus postcranial material) should provide new
information that would either strengthen or require an emendation of the interrelationships of the
three edaphosaurid genera given here.

Caseidae

Eothyris

Mycterosaurus

Ophiacodontidae

Haptodus

Ianthasaurus

Glaucosaurus

Edaphosaurus

text-fig. 3. Cladogram illustrating a hypothesis of relationships for Glaucosaurus megalops. Outgroup taxa
used in the analysis are not shown. Nodes are defined by the following unambiguous characters: a (defining
Synapsida), septomaxilla vertically oriented with medial process, lateral temporal opening present, occiput
slopes anterodorsally. b (defining Caseasauria), frontal contribution to orbital margin reduced, c (defining
Eupelycosauria), antorbital region tall, supratemporal narrow, parietal foramen positioned posteriorly on
interparietal suture, d, jugal separates maxilla and quadratojugal, posterior cheek margin concave, stapes
articulates in pocket on opisthotic, angular with ventral keel, e, quadratojugal reduced in size, prearticular
twisted, f (defining Edaphosauridae), presacral neural spines with lateral tubercles, presacral neural spines
subcircular in cross section, anterior presacral neural spines lean anteriorly, posterior presacral neural spines
curve posteriorly, g, caniniform region absent, caniniform tooth absent, prefrontal ventral process expanded

medially.

The addition of Glaucosaurus to the Edaphosauridae provides valuable insight into the evolution
of herbivory in edaphosaurids. As the sister taxon of Edaphosaurus , Glaucosaurus features several
characters which have been attributed to the adaptation to herbivory in Edaphosaurus (Modesto
1991). These include the presence of an antorbital buttress, isodonty, loss of the transverse flange
of the pterygoid, and perhaps also the presence of an abbreviated snout. Ancestrally, edaphosaurids
may have fed upon soft-bodied arthropods, a diet that has been proposed for Ianthasaurus (Reisz
and Berman 1986). The ancestor of Glaucosaurus and Edaphosaurus may have progressed to a
(possibly omnivorous) diet of harder foods; this is suggested by the short, buttressed antorbital
region and isodonty present in both daughter taxa. Glaucosaurus may have been specialized for
feeding upon hard-bodied arthropods, judging from the presence of sharp, laterally-compressed
teeth, whereas Edaphosaurus evolved true herbivory.

Acknowledgements. I thank Dr John R. Bolt, of the Chicago Field Museum of Natural History, for the loan
of the holotypes of Glaucosaurus and Edaphosaurus novomexicanus. I also thank Dr Farish Jenkins and Mr
Chuck Schaff, of the Harvard Museum of Comparative Zoology, for the loan of additional specimens. I am
grateful to Mr Michael deBraga, for procuring the type specimen on my behalf, and for his criticisms of the
typescript. I am also indebted to Dr Robert R. Reisz, for discussion and suggestions, and to Mr Michel Laurin,

58

PALAEONTOLOGY, VOLUME 37

for helpful commentary and discussion, and for the use of his computer for the phylogenetic component of the
paper.

REFERENCES

berman, d. s and reisz, r. r. 1982. Restudy of Mycterosaurus longiceps (Reptilia, Pelycosauria) from the
Lower Permian of Texas. Annals of the Carnegie Museum , 51, 423-453.

— and eberth, D. a. 1987. Seymouria sanjuanensis (Amphibia, Batrachosauria) from the Lower
Permian Cutler Formation of north-central New Mexico and the occurrence of sexual dimorphism in that
genus questioned. Canadian Journal of Earth Sciences , 24, 1769-1784.

brinkman, d. and eberth, d. a. 1983. The interrelationships of pelycosaurs. Breviora , 473, 1-35.
broom, R. 1932. The mammal-like reptiles of South Africa. H. F. and G. Witherby, London. 376 pp.
cope, e. d. 1882. Third contribution to the history of the Vertebrata of the Permian Formation of Texas.

Proceedings of the American Philosophical Society , 20, 447^46 1 .
dilkes, d. w. and reisz, r. r. 1986. The axial skeleton of the Early Permian reptile Eocaptorhinus laticeps
(Williston). Canadian Journal of Earth Sciences , 23, 1288-1296.

Gauthier, j. a., kluge, a. g. and rowe, t. 1988. The early evolution of the Amniota. 103-155. In benton, m. j.
(ed.). The phylogeny and classification of the tetrapods. Volume 1, Amphibians , reptiles , birds. Systematics
Association, Special Volume 35a, Clarendon Press, Oxford, 377 pp.
heaton, M. J. 1979. Cranial anatomy of primitive captorhinid reptiles from the Late Pennsylvanian and Early
Permian, Oklahoma and Texas. Oklahoma Geological Survey, Bulletin 127, 1-84.
hentz, T. F. 1988. Lithostratigraphy and paleoenvironments of Upper Paleozoic continental red beds, North-
Central Texas: Bowie (new) and Wichita (revised) Groups. University of Texas at Austin, Bureau of
Economic Geology Report of Investigations, 170, 1-55.

holmes, r. B. 1977. The osteology and musculature of the pectoral limb of small captorhinids. Journal of
Morphology, 152, 101-140.

kemp, t. s. 1982. Mammal-like reptiles and the origin of mammals. Academic Press, London, 363 pp.
laurin, m. 1993. Anatomy and relationships of Haptodus garnettensis, a Pennsylvanian synapsid from Kansas.
Journal of Vertebrate Paleontology, 13, 200-229.

lewis, G. E. and vaughn, p. p. 1965. Early Permian vertebrates from the Cutler Formation of the Placerville
area, Colorado. US Geological Survey, Professional Paper 503-C, 1-50.
modesto, s. p. 1991. Cranial anatomy of advanced edaphosaurid synapsids from the Lower Permian of New
Mexico and Texas. Unpublished M.Sc. Thesis, University of Toronto.

— and reisz, R. r. 1990. A new skeleton of Ianthasaurus hardestii, a primitive edaphosaur (Synapsida:
Pelycosauria) from the Upper Pennsylvanian of Kansas. Canadian Journal of Earth Sciences, 26, 834-844.

1992. Restudy of Permo-Carboniferous synapsid Edapliosaurus novomexicanus , the oldest known
herbivorous amniote. Canadian Journal of Earth Sciences, 29, 2653-2662.
osborn, h. f. 1903. On the primary division of the Reptilia into two subclasses, Synapsida and Diapsida.
Science. 17, 275-276.

reisz, r. r. 1986. Pelycosauria. In wellnhofer, p. (ed.). Handbuch der Paldoherpetologie, Teil 17A. Gustav
Fischer Verlag, Stuttgart. 102 pp.

— and berman, D. s 1986. Ianthasaurus hardestii n. sp., a primitive edaphosaur (Reptilia, Pelycosauria) from
the Upper Pennsylvanian Rock Lake Shale near Garnett, Kansas. Canadian Journal of Earth Sciences, 23,
77-91.

— and scott, d. 1992. The cranial anatomy and relationships of Secodontosaurus, an unusual
mammal-like reptile (Synapsida: Sphenacodontidae) from the early Permian of Texas. Zoological Journal of
the Linnean Society, 104, 127-184.

— and dilkes, d. w. 1992. The taxonomic position of Anningia megalops, a small amniote from the Permian
of South Africa. Canadian Journal of Earth Sciences, 29, 1605-1608.

romer, a. s. and price, l. i. 1940. Review of the Pelycosauria. Geological Society of America, Special Paper, 28,
1-538.

sigogneau-russell, d. and russell, d. e. 1974. Etude du premier Caseide (Reptilia, Pelycosauria) d’Europe
occidentale. Bulletin du Museum National d'Histoire Naturelle, 3e Serie, 230, 145-216.
stovall, j., price, l. i. and romer, a. s. 1966. The postcranial skeleton of the giant Permian pelycosaur
Cotylorhynchus romeri. Bulletin of the Museum of Comparative Zoology, 135, 1-30.
swofford, d. 1989. PAUP: Phylogenetic Analysis Using Parsimony, Version 3.0. Privately circulated
manuscript of the instruction manual.

MODESTO: LOWER PERMIAN SYNAPSID

59

white, T. E. 1939. Osteology of Seymouria havlorensis. Bulletin of the Museum of Comparative Zoology , 85,
323-410.

williston, s. w. 1915. New genera of Permian reptiles. American Journal of Science , 39, 575-579.
wilson, h. 1989. Ophiacodon beckiae: A new species of pelycosaur of the family Ophiacodontidae (Reptilia,
Synapsida), from the Late Pennsylvanian of Kansas. Unpublished M.Sc. Thesis, University of Toronto.

s. P. MODESTO

Department of Zoology
Erindale College
University of Toronto

Typescript received 27 April 1993 Mississauga

Revised typescript received 16 August 1993 Ontario, Canada L5L 1C6

ABBREVIATIONS USED IN THE TEXT-FIGURES

a

angular

n

nasal

pra

prearticular

ac

anterior coronoid

P

parietal

prf

prefrontal

d

dentary

pal

palatine

Pt

pterygoid

f

frontal

pc

posterior coronoid

qj

quadratojugal

J

jugal

po

postorbital

sm

septomaxilla

1

lacrimal

pm

premaxilla

sp

splemal

m

maxilla

PP

postparietal

v.pr.prf

ventral process of prefrontal

APPENDIX 1

Description of the characters used in the phylogenetic analysis. Characters are listed in order of their
location on the skull, the mandible, and the postcranial skeleton.

1. Marginal teeth: taper gradually (0) or are slightly bulbous (1).

2. Premaxillary dentition: first tooth equal to or smaller than (0) or larger than maxillary teeth (1) in basal
cross-section.

3. Caniniform region: present (0) or absent (1).

4. Caniniform tooth: absent (0) or present (1).

5. Maxillary teeth: subcorneal (0) or compressed laterally (1).

6. Maxilla: contacts (0) or separated from (1) quadratojugal.

7. Maxilla: short, does not extend posterior to posterior orbital margin (0) or long, extending past orbit (1).

8. Septomaxilla: sheet-like and curved (0) or oriented vertically with medial flange (1).

9. Antorbital region: low (0) or tall (1).

10. Prefrontal ventral process: laterally compressed (0) or expanded transversely (1).

1 1. Frontal: bounded by pre- and postfrontal laterally (0), separates pre- and postfrontals (1) and with broad
orbital exposure (2).

12. Supraorbital margin: weakly developed (0) or expanded laterally (1).

13. Parietal foramen: positioned anteriorly (0), at midpoint of (1), or posteriorly (2) on interparietal suture.

14. Supra temporal : large, broad (0) or long, narrow (1).

15. Quadratojugal : large and contributes to subtemporal bar (0) or small and covered laterally by squamosal

(1).

16. Lateral temporal fenestra: absent (0) or present (1).

17. Posterior cheek margin: straight (0) or concave (1).

18. Skull: long, eight dorsal centra or more in length (0), or short, six dorsal centra or less in length (1).

19. Postorbital region: shorter than (0) or equal to or longer than (1) antorbital region.

20. Occiput: vertical (0) or slopes anterodorsally (1).

21 Stapes: dorsal process free (0) or articulates in pocket on opisthotic (1).

22. Pterygoid: transverse flange present (0) or absent (1).

23. Prearticular : straight (0) or twisted posteriorly (1).

24. Angular: ventral keel absent (0) or present (1).

25. Cervical centra: equal to or longer than (0) or shorter than (1) mid-dorsal centra.

26. Presacral neural spines: laterally compressed (0) or subcircular (1) in distal cross section.

60

PALAEONTOLOGY, VOLUME 37

27. Presacral neural spines: anterior spines extend dorsally (0) or lean anteriorly (1).

28. Presacral neural spines: posterior spines extend dorsally (0) or curve posteriorly (1).

29. Presacral neural spines: lateral tubercles absent (0) or present (1).

30. Neural arches: not excavated (0) or excavated (1).

31. Dorsal vertebrae: transverse processes moderately developed (0) or elongate (1).

32. Dorsal ribs: curved proximally (0) or curved throughout length (1).

33. Dorsal ribs: tubercula well developed (0) or greatly reduced (1).

34. Ilium: anterodorsal process absent (0), moderately developed (1), or strongly developed (2).

APPENDIX 2

Distribution of the character states among the taxa examined in the analysis. The numbers in the
top column (1-34) refer to the characters described in Appendix 1. A question mark indicates that

the character state could not be determined because of missing data.
Character number 1 2 3 4 5 6 7 8 9 1011

12

13

14

15

16

17

Taxon

Seymouria

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Captorhinus

0

1

0

1

0

1

0

0

0

1

2

0

0

0

0

0

0

Caseidae

1

1

1

0

0

0

1

1

0

0

1

0

0

0

0

1

0

Eothyris

0

0

0

1

0

0

1

7

0

1

1

0

1

0

0

1

0

Mycterosaurus

0

1

0

1

1

0

1

7

1

0

2

0

2

1

0

1

0

Ophiacodontidae

0

1

0

1

0

1

0

7

1

0

2

0

2

1

0

1

1

Haptodus

0

1

0

1

1

1

0

1

1

0

2

0

2

1

1

1

1

Glaucosaurus

0

0

1

0

1

1

1

1

1

1

?

0

?

7

?

1

1

Ianthasaurus

0

9

0

1

0

1

0

7

1

0

2

0

1

1

1

1

1

Edaphosaurus

1

0

1

0

0

1

0

1

1

1

1

1

1

1

1

1

1

Character number

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

Taxon

Seymouria

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Captorhinus

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Caseidae

1

0

1

0

0

0

0

1

0

0

0

0

0

1

1

2

2

Eothyris

?

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

Mycterosaurus

0

1

1

0

0

0

0

0

0

0

0

0

1

0

0

0

0

Ophiacodontidae

0

1

1

1

0

0

1

0

0

0

0

0

0

0

0

0

0

Haptodus

0

1

1

1

0

1

1

0

0

0

0

0

1

0

0

0

0

Glaucosaurus

?

?

7

7

1

?

?

?

?

?

?

7

7

?

?

?

?

Ianthasaurus

0

1

1

7

7

1

1

0

1

1

1

1

1

0

0

0

1

Edaphosaurus

1

0

1

1

1

1

1

1

I

1

1

1

0

1

1

2

2

CEPHALASPIDS FROM THE LOWER DEVONIAN OF
PRINCE OF WALES ISLAND, CANADA

by D. L. DINELEY

Abstract. A Lochkovian vertebrate assemblage from the Peel Sound Formation of northern Prince of
Wales Island, Northwest Territories, Canada, includes numerous headshields resembling the osteostracan
Parameteoraspis , together with several fragments suggesting Machairaspis. Five specimens of the former,
distinct in their extended and diverging pectoral cornua, may be regarded as a conspecific variety. A possible
advantage afforded by such cornua, in this and other species, may be their use in swimming and in disturbing
the substrate during feeding. A similar role for the rostra of other osteostracans and eugaleaspids is suspected.

The Silurian and Devonian vertebrate-bearing rocks of Arctic Canada lie within Young’s (1981)
Euramerican or Cephalaspid vertebrate province of the Lower Devonian. Cephalaspids, as the
distinguishing element in the faunas, occur widely throughout the province, and constitute a diverse
but puzzling group. They are conspicuous though not abundant in the Downtonian and Dittonian
Series in the British Isles and Western Europe (Stensio 1932; Heintz 1940; White 1961, 1985),
Svalbard (Wangsjo 1952; Janvier 1985a) and Podolia (Janvier 19856, 1988). Osteostraci appear to
have little stratigraphical value in much of the Euramerican Province. Only in Spitsbergen has it
been possible to plot their stratigraphical ranges on the basis of common, if not abundant, material
within the Red Bay and Wood Bay formations (Janvier 1985a), which are Pndoli to Lochkov in age.

On the North American mainland they are relatively less abundant but are known from the
Atlantic area (Robertson 1936; Dineley 1967; Pageau 1969) and from several places in the
southwestern States (Denison 1952). From northern Canada they have been recorded from a very
few localities on Somerset Island (Dineley 1968) and from the MacKenzie Mountains (Dineley and
Loeffler 1976). In all these vertebrate ‘communities’ the Osteostraci occupy a comparatively minor
role, being neither abundant nor taxonomically diverse in comparison with the other members. The
material described here, however, comes from a small volume of rock in which the cephalaspids are
relatively common within a very restricted assemblage or faunule, which presumably may represent
a similar ‘community’. The state of preservation is poor, making taxonomic identification difficult,
but this does not obscure the wide variation in the shape of the headshield in what is here regarded
as a single species.

LOCALITY AND FAUNA

The cephalaspid-yielding horizon on Prince of Wales Island lies relatively high in the
sandstone-carbonate facies of the upper member of the local Peel Sound Formation. This unit is
essentially confined to the vicinity of the Boothia Uplift where the Silurian to Devonian succession
exhibits proximal Old Red Sandstone facies passing laterally into distal marine carbonates and
minor elastics (Miall 1970). Stewart (1987) has given an overall account of the stratigraphy of the
region, noting the occurrence of vertebrates in the marine facies of the Ludlovian and Pfidoli and
also in the non-marine rocks of the Peel Sound Formation. Dineley (1990) has outlined the
vertebrate palaeontology, drawing attention to the scattered but very numerous localities at which
vertebrate remains occur. On the basis of both agnathan and gnathostome vertebrates and the
invertebrates, Elliott (1984) broadly correlated the horizon from which the present fauna comes

| Palaeontology, Vol. 37, Part I, 1994, pp. 61-70, 1 pl.|

© The Palaeontological Association

62

PALAEONTOLOGY, VOLUME 37

text-fig. 1 a, part of Northwest Territories, Canada, with the position of Text-figure Ib indicated, b, the
cephalaspid-yielding locality (F) in northern Prince of Wales Island.

with the vogti Horizon in Spitsbergen, with the lower part of the crouchi Zone in Britain and with
the Czortkow Horizon in Podolia. This is within the Lochkovian Stage.

The locality is on the east bank of a small water-course near the north coast (the Baring Channel)
of Prince of Wales Island (F on Text-fig. 1, and given as locality A in Dineley 1976, p. 2). Here a
green-weathering fine sandstone contains:

Ctenaspis obruchevi Dineley common

Cyathaspid indet. cf. Pionaspis rare

? Baringaspis dineley i Miles very rare

? Parameteoraspis cf. P. oblonga (Wangsjo) common

Cephalaspid indet. cf. Machairaspis sp. very rare

Virtually all the fossils are fragmentary or isolated headshields, portions of cephalaspid, and
other squamation and minor indeterminate debris. Several Ctenaspis and cephalaspid headshields
bear attached squamation; the former include complete trunk and tail parts (Dineley 1976). The
dorsal and ventral Ctenaspis shields are separated and the scales seem to be attached to the dorsal
shields only. These remains have been water-transported, presumably soon after the death of the
animals and before complete disintegration took place. As is the case in virtually all cephalaspid
occurrences, the headshields appear to be the right way up. No marine invertebrate fossils
accompany the vertebrates and a freshwater or brackish environment of deposition is possible.
There are no microvertebrates. The assemblage seems to represent a single benthonic, if not
infaunal, community, together with the active nektonic arthrodire Baringaspis. It is unique amongst
the many vertebrate-bearing levels in the vicinity of the Boothia Uplift in the relatively large number
of cephalaspid individuals present: in all other assemblages Osteostraci are very rare.

SYSTEMATIC PALAEONTOLOGY

Order osteostraci Lankester, 1868
Suborder cornuata Janvier, 1985a
Family Group cephalaspidiens Janvier, 1985a
Genus parameteoraspis Blieck et a /., 1987

Type species. Cephalaspis gigas Wangsjo, 1952.

D1NELEY: CEPHALASPIDS FROM PRINCE OF WALES ISLAND

63

table 1. Measurements (in mm) of cephalaspid headshields from the locality near Baring Channel, Prince of
Wales Island, a, length of median from rostral margin to level of cornual tips; b, distance between cornual tips;
c, distance between median point of rostral margin and cornual tip; *, form with elongate and divergent
cornua. Number of specimens measured = 21.

Spec. No.

A

B

C

A/B

12471

48

52

—

0-92

12474

—

62

56

—

12476

75

55

—

1 36

12477*

—

67

72

—

12478

—

59

61

—

12480

50

70

—

0-71

12481

—

62

62

—

12482

55

64

62

0-85

12483

42

65

64

0-64

12484*

—

70

72

—

12487

—

70

66

—

12488

55

60

61

0 91

12489

50

70

66

0 71

12496

58

65

64

0-89

12497

—

65

64

—

12498

—

69

52

—

12501

52

62

—

0-83

12503

40

60

—

0-66

12504

—

72

68

—

12506*

64

77

82

0-83

1251 1

58

69

66

0-60

Average

(typical)

53

60

68

—

Mean

Range

57-5

61

60

0-60-0-92

Parameteoraspis cf. P. oblonga (Wangsjo, 1952)

Plate 1, figures 1-6, Text-figure 2a-e

Material. Fragmentary head shields, some with adjacent squamation; Canadian Museum of Nature, Ottawa,
NMC 12471-12498, NMC 12501-12506, 12511.

Locality and horizon. Stream bank near north coast of Prince of Wales Island, arctic Canada; Lochkovian,
Lower Devonian.

Description. The bulk of the osteostracan material (PI. 1 ; Text-fig. 2a-e) appears to belong to a medium-size
true cephalaspid, tentatively compared below to Parameteoraspis oblonga (Stensio). In addition, two other
fragments of cephalaspid head shield bear relatively large orbits and traces of a large medial spinal crest such
as is found in Machairaspis.

Some twenty-five specimens preserve most of the outline of a cephalaspid head shield in dorsal or ventral
aspect. While the outline of each is intact, it has not been possible to distinguish much of the detailed
morphology, such as sensory fields, nerve or vascular canals, or the positions of the nasal, pineal or orbital
apertures. The bone itself is very thin and has been severely crushed. Additionally, vertical compression of the
head shield has deformed the median profile. No external ornamentation has been observed.

The dimensions are given in Table 1 . The size range is small and the animals were presumably all of the same
generation or approximate age. Size was regarded as a specific character by Stensio (1932) and by Wangsjo
(1952). White (1985), however, noted as much as two hundred per cent variation in some British species.

64

PALAEONTOLOGY, VOLUME 37

The fossils represent a group of medium-size cephalaspidids sensu stricto (Cephalaspidiens of Janvier 1985 a)
with very prominent pectoral cornua that lack marginal denticles. The rim of the headshield varies little in
shape, except in five individuals. The majority exhibit a smooth parabolic curve from the rostral median point
to the cornua. There is no rostral lip, process or thickening of the rim. The pectoral cornua are broad and
directed posteriorly or postero-laterally. They appear to have been originally thin or flat and subsequently
further dorso-ventrally compressed. The pectoral sinus is rather narrow to broad and the interzonal part of the
shield extends backwards to almost half the length of the cornua. As is normal in the cephalaspids, the ventral
surface of the head is flat; the dorsal is domed to a height of about 15 mm in the posterior median part but
with wide flat margins. In plan view, the general outline with broad-based cornual processes is similar to
Cephalaspis lyelli Agassiz (White 1958).

Such evidence as there is of the original dorsal surface suggests that the lateral fields were probably
inconspicuous, long and narrow, extending well into the cornual regions. The size, position and shape of the
central sensory field are obscure, but the orbits are small, situated high, about halfway back from the anterior
margin and just in front of the presumed position of the central field. Neither nasal nor pineal apertures have
been identified. Poor preservation also precludes the recognition of dorsal surface ornamentation, nor has any
trace of the features of the ventral surface survived (Plate 1). Thus it has not been possible to determine if
tuberculation (a diagnostic character) was originally present. Several specimens (NMC 12476, 12489, 12500,
12504, 12506b, for example) retain impressions or fragments of the scale rings immediately behind the
headshield. The scales are similar to those of other cephalaspids, being narrow and tall, with appreciable
overlap posteriorly. So far, no vestige of a large median dorsal scute, as present in the true cephalaspidians,
has been found. No sign of pectoral appendages is seen.

Five head shields (NMC 12476, 12482, 12484, 12497, 12506) differ from the majority in dorsal view, having
broadly divergent cornua (Text-fig. 2c). The median length is about 45 mm but the points of the cornua are
some 80 mm apart. The anterior margins of these specimens are broadly rounded with no apical angle or
rostral feature, the lateral margins are straight almost to the hind tip. Thus the cornua are directed
posterolaterally, as in such species as Parameteoraspis caroli (Wangsjo). The pectoral sinus is wide and short.
P. caroli is a larger cephalaspid, being some 60 by 18 mm in size. As far as can be seen, all other features in
these five specimens correspond to those in the remainder. Such a marked difference in outline of the head
shield has been used as a specific character by previous authors, and at least four such species have been
described from the Wood Bay Formation of Spitsbergen (Janvier 1985 a) where there are big size differences
between the species. It is not proposed to raise the present specimens to higher taxonomic rank than variety
here.

On the meagre evidence from this material there is no obvious taxon to which it may be referred. It
nevertheless seems to be close to the species Parameteoraspis oblonga (Stensio). The genus contains species both
larger and smaller in size than that of the present specimens. It is now recorded also from the Lower Emsian
‘Klerfer Schichten’ of the Rheinisches Schiefergebirge (Bardenheuer and Janvier, 1990).

Discussion. Janvier ( 1981 ; 1985a p. 200) defined Parameteoraspis (under its previous invalid name
Meteoraspis ) thus;

Cephalaspidians of generally large size; cornual processes long and proximally wide, bearing the
posterior part of the lateral fields, which is narrow and follows the lateral borders of the cornua almost
to their ends. The median dorsal field is also wide and short, including the external openings of the
endolymphatic ducts. The pineal plate is replaced by separate tesserae of the same kind as those
covering the median dorsal field. The naso-hypophysial opening is an elongated slit. The superficial
layer of the exoskeleton is continuous in mature individuals, but is smooth except around the orbits
and along the margins of the shield, where it carries very small irregular tubercles.

His (emended) diagnosis of Wangsjo’s (1952) species M. oblonga (Stensio) reads;

Meteoraspis of medium dimensions, shield longer than broad, with parabolic rostral margin and
cornual processes directed posteriorly. Spinal crest present but projecting only slightly.

EXPLANATION OF PLATE 1

Figs 1-4. ? Parameteoraspis cf. oblonga (Stensio); Figs 5-6. ? Parameteoraspis cf. oblonga var. All from the Peel
Sound Formation of northern Prince of Wales Island, Canada. 1, NMC 12482a; 2, NMC 12476; 3, NMC
125066; 4, NMC 12483; 5, NMC 12488; 6, NMC 12489. All figures c. x 1 .

PLATE 1

DINELEY, 1 PARA METEOR A SPIS

text-fig. 2. ? Parameteoraspis cf. oblonga (Stensio) from the Upper Member of the Peel Sound Formation,
northern Prince of Wales Island, Canada; poorly preserved cephalic shields with traces of squamation.
a, NMC 12483; b, NMC 12489; c, NMC 12474; d, NMC 12487; e, NMC 12506. All figures c.xO-6.

There is little new material to amplify the original diagnoses, which nevertheless approximate more
closely than others to the present specimens (Text-fig. 3). The Prince of Wales Island material shows
no trace of tubercles on the marginal surface nor of a spinal crest.

The five specimens with laterally extended cornua conform generally with these criteria except for
the diverging axes of these features. Their appearance in an otherwise uniform sample of a species
of cephalaspid suggests that they may embody a response to a particular environmental
opportunity. It may follow that should the opportunity have persisted the response may have led
to the establishment of a separate species of which these few individuals were possibly the
forerunners. So far no further cephalaspids have been found at higher levels in this facies, nor have
such distinctively cornuate forms been found elsewhere in this region.

The four or more Spitsbergen species with widely divergent pectoral cornua have been described
from rather sparse material. Their characteristics generally, however, are distinct, whereas the
Canadian specimens are distinguished from the more numerous individuals present only by their
pectoral width.

Genus machairaspis Janvier 1985u

Type species. Cephalaspis corystis (Wangsjo, 1952)

DINELEY: CEPHALASPIDS FROM PRINCE OF WALES ISLAND

67

text-fig. 3. a, Parameteoraspis oblonga (Stensio). b, Parameteoraspis caroli (Wangsjo). Both figures are alter

Janvier (1985). Scale bar represents 10 mm.

text-fig. 4. Cephalaspid indet. cf. Machairaspis. a-b, NMC 12517; left lateral and dorsal views; c. x 1.

cf. Machairaspis sp.

Text-figure 4a-b.

Material. Two fragments of cephalaspidian head shield bearing the basal part of a median dorsal spine or vane
(NMC 12517, 12534).

Locality and horizon. As for Parameteoraspis cf. P. oblonga. In Spitsbergen Machairaspis ranges throughout
the Fraenkelryggen and Ben Nevis Formations into the Kapp Kjeldsen Division of the Wood Bay Formation;
it therefore extends from the Lochkovian into the lower Pragian.

Description. The presence of fragments of osteostracan headshields with a median spinal process indicates the
presence of another genus. Genera bearing a dorsal spine or vane are few and rare. The zenaspidians
(scolenaspidians of Janvier 1985 a, p. 116) Scolenaspis and Machairaspis are the prime examples. The material
from Prince of Wales Island represents a form somewhat larger than Parameteoraspis , but generally small
compared with the Spitsbergen species. No surface ornament or tuberculation is apparent. The orbits are large

68

PALAEONTOLOGY, VOLUME 37

and set high on the shield; the leading edge of the spinal process is strongly inclined to the posterior, but
specimen NMC 12517 is deformed and identification as Machairaspis is most tentative.

Discussion. Although the variety of wide headshield is represented by only five specimens, they are
of constant proportions and do not seem to be deformed or misshapen in any significant way. They
may represent a form that may have been at no disadvantage in competition with the forma typica.

The headshield itself is an almost complete rigid box of bone; only its floor between mouth,
branchial openings and the abdominal exoskeleton, was in any way flexible. This, and the
mouthparts, were required to move during feeding and respiration. The general structure of
cephalaspids is consistent with a benthonic mode of life wherein the development of pectoral glide
planes may have assisted in lift-off from the bottom when required for feeding and breathing and
especially during locomotion. Pectoral appendages would effectively provide additional and
directional drive. Belles-Isles (1987) has analysed swimming modes and hydrodynamics for two
agnathan species, Alaspis macrotuberculata and Pteraspis rostrata.

text-fig. 5. Late Silurian and Early Devonian Agnatha with both long pectoral cornua and long rostra, a,
Lungmenshcinaspis ; B, benneviaspid ; c. Boreas pis ; d, Sanqiaspis ; E, Asiaspis; F, Spatulaspis', G, Dicrcinaspis.
a, e, G, from South China; b-d, f, from Spitsbergen. Figures not drawn to scale.

The late Devonian osteostracan Alaspis is morphologically similar to Parameteoraspis, and
although no pectoral appendages have yet been discovered in the latter, it may be assumed that the
two species functioned in similar fashion. Janvier (1985a, p. 225) and Belles-Isles (1987) considered
that in the Boreaspidae the extended rostral and pectoral processes may have had a defensive
function against predators. It is, however, hard to see how extended cornua would have given a
defensive advantage. Their hydrodynamic function is still not fully understood. Belles-Isles (1987,
p. 367) emphasized the role of cornua in increasing manoeuvrability, and noted that the
environment was frequently turbulent. Janvier’s (1985a) cladistic analysis of the Spitsbergen
osteostraci draws a fundamental division between cornuate and non-cornuate taxa. The latter are

DINELEY: CEPHALASPIDS FROM PRINCE OF WALES ISLAND

69

the earlier (Silurian) forms; the cornuates include his groups cephalaspidiens, thyestidiens,
scolenaspidiens, kiaeraspidiens and benneviaspidiens, of which only the thyestidiens are primarily
Silurian. All the Devonian forms are essentially components of the Old Red Sandstone (i.e. non-
marine) vertebrate faunas, and were adapted for life in the vigorous, if not highly turbulent, waters
of that realm. The development of strong cornual processes was most probably related to life in such
ca habitat (Janvier 1985 c). Extended rostra and cornual processes also occur in certain Chinese
eugaleaspids (e.g. Lungmenshanaspis , see Text-fig. 5). As mentioned above, Janvier (1985 a, p. 225),
like Belles-Isles (1987), considered that in the Boreaspididae the rostral and pectoral processes may
have had a dissuasive function against predators such as porolepiforms or arthrodires. The purpose
of these long cephalic features is, nevertheless, thought here to be associated not only with
locomotion but also with probing or disturbing sediment and vegetation in the search for food.

Osteostracan squamation leads to the view that the trunk and tail were capable of sinuous lateral
flexure to a considerable degree, the rigid nature of the scales notwithstanding. Flexure in the
vertical plane, however, was probably very limited. The heterocercal tail would have functioned as
a large vane moving from side to side to provide a thrust at right angles to itself and contributed
a measure of lift. The pectoral fins, set at an angle of incidence to the axis of movement would have
carried much of the load in swimming, just as does an aeroplane wing. They would also, as noted
above, assist in steering and variable horizontal movement. Thus the cephalaspids, if not all
osteostracans, swam by anguilliform motion of the body whereby a wave (movement) progresses
backwards through the body to the tail. The speed of the wave along the body is always greater than
the speed of the fish through the water. The amplitude of the wave increases as it moves back from
head to tail. While the function of the pectoral appendages would have helped to maintain course
or steer they would also serve to move the head in a broad sweeping movement from side to side
(Belles-Isles 1987, fig. 7). Thus anguilliform movement could be initiated by movement of the head
shield, and at the sediment-water interface would disturb organisms and food particles which could
then be drawn into the mouth. A rostrum would have effectively disturbed sediment or vegetation
forward of the mouth, i.e. normally perhaps upstream, and current activity would move food
particles towards the mouth. The disruption of thin algal mats could have been effected by even a
small rostrum; the larger rostra could probe filamentous or vascular plant thickets. Examples of
rostra capable of this kind of activity are included in Text-figure 5. The question arises as to whether
these features be considered as ‘advanced’ specializations which fitted the animals better to their
particular habitat. Long rostra and pectoral cornua arose, as noted above, in several different
groups of Devonian agnathans, all associated with the clastic sedimentary facies. Perhaps the
possession of these structures offered an advantage such as greater stability when at rest in a habitat
founded upon unconsolidated sandy substrates and where frequent shifting of food sites occurred.

It would be interesting to discover if a higher proportion of the cephalaspids in later communities
within this facies at Baring Channel possessed extended cornua. The ‘community’ described here
is singular in the relatively large number of cephalaspid individuals it contains. Ctenaspis also is
common here and appears probably to have had a largely infaunal existence as a detritus feeder.
The other vertebrates may also have spent much of their lives as benthos, all being subject to
periodic episodes of turbulence and of violent current activity in the water above them.

Acknowledgements. Dr E. J. Loeffler kindly read and suggested improvements to an early draft of this paper.
Dr Ph. Janvier also provided welcome critical assistance and discussion. The photographs were taken by Simon
Powell; J. Andrew Lawrence (Keele University) drafted Text-figure 1.

REFERENCES

bardenheuer, p. von and janvier, p. 1990. Cephalaspiden (Osteostraci) aus dem Unterdevon (Emsium) von
Waxweiler (Rheinishes Schiefergebirge, Deutschland). Neues Jahrbuch Jiir Geologie und Palaontologie
Monatshefte , 1990. 639-646.

belles-isles, m. 1987. La nage et 1’hydrodynamique de deux Agnathes du Paleozoique: Alaspis macro-
tuberculata et Pteraspis rostrata. Neues Jahrbuch fur Geologie und Palaontologie , Abhandlungen , 175,
347-376.

70

PALAEONTOLOGY, VOLUME 37

blieck, a., goujet, d. and janvier, p. 1987. The vertebrate stratigraphy of the Lower Devonian (Red Bay
Group and Wood Bay Formation) of Spitsbergen. Modern Geology , 11, 197-217.
denison, r. l. 1952. Early Devonian fishes from Utah. Part I. Osteostraci. Fieldiana , Geology , 11, 198-218.
dineley, D. L. 1967. The Lower Devonian Knoydart Fauna. Journal of the Linnean Society of London , 47,
1 5-29.

— 1968. Osteostraci from Somerset Island. Bulletin of the Geological Survey of Canada , 165, 249-263.
1976. New species of Ctenaspis (Ostracodermi) from the Devonian of Arctic Canada. 26-44.

In churcher, c. s. (ed.). Athlon: essays in palaeontology in honour of Loris Shano Russell. Royal Ontario
Museum, Toronto, 286 pp.

- 1990. Palaeozoic fishing the Franklinian Grounds. 55-80. In harington, c. r. (ed.). Canada's missing
dimension , 1. Canadian Museum of Nature, Ottawa, 834 pp.

— and loeffler, e. j. 1976. Ostracoderm faunas of the Delorme and associated Siluro-Devonian formations.
Northwest Territories, Canada. Special Papers in Palaeontology, 18, 1-214.

elliott, D. K. 1984. Siluro-Devonian fish biostratigraphy of the Canadian Arctic Islands. Proceedings of the
Linnean Society of New South Wales , 107, 197-209.

heintz, a. 1940. Cephalaspida from the Downtonian of Norway. Skrifter, Norske Vidensk-Akademi
mathematik-naturhistoriska Klasse , Oslo , 1939 (5), 1-119.
janvier, p. 1981. Norse/aspis glacialis n.g., n.sp. et les relations phylogenetiques entre les Kiaeraspidiens
(Osteostraci) du Devonien inferieur de Spitsberg. Palaeovertebrata, II, 19-131.

1985a. Les Cephalaspides du Spitsberg. C aiders de Paleontologie (section vertebres), Centre National de
Recherches Scientifiques, 244 pp.

1 985/?. Preliminary description of Lower Devonian Osteostraci from Podolia (Ukrainian SSR). Bulletin
of the British Museum ( Natural History ), Geology Series, 38, 309-334.

1985c. Environmental framework of the diversification of the Osteostraci during the Silurian and
Devonian. Philosophical Transactions of the Royal Society of London , Series B, 309, 259-272.

1988. Un nouveau cephalaspide (Osteostraci) du Devonien inferieur de Podolie (R.S.S. d'Ukraine). Acta
Palaeontologica Polonica, 33, 353-358.

miall, a. d. 1970. Continental marine transition in the Devonian of Prince of Wales Island, Northwest
Territories. Canadian Journal of Earth Sciences, 7, 125-144.
pageau, y. 1969. Nouvelle faune ichthyologique du Devonien moyen dans les Gres de Gaspe (Quebec). 2,
Morphologie et systematique. Premier section: A, Eurypterides; B, Ostracodermes; C, Acanthodiens et
Selachiens. Naturalist Canadien , 96, 399-478.

robertson, G. m. 1936. New cephalaspids from Canada. American Journal of Science, 31, 288-295.
stensio, e. a. 1927. The Downtownian and Devonian vertebrates of Spitsbergen. 1. Family Cephalaspidae.
Skrifter om Svalbard og Ishavet, 12, 1-391.

1932. The cephalaspids of Great Britain. British Museum (Natural History) London, 220 pp.
stewart, w. d. 1987. Late Proterozoic to Early Tertiary stratigraphy of Somerset Island and northern Boothia
Peninsula, District of Franklin, N.W.T. Geological Survey of Canada, Paper 83-26, 1-78.
wangsjo, G. 1952. The Downtownian and Devonian vertebrates of Spitsbergen. IX. Morphologic and
systematic studies of the Spitsbergen cephalaspids. Results of Th. Vogt’s expedition 1928 and the
English-Norwegian-Swedish expedition 1939. Skrifter om Svalbard og Ishavet, Oslo , 12, 1-391.
white, e. i. 1958. On Cephalaspis lyelli Agassiz. Palaeontology, 1, 99-105.

- 1961 . The Old Red Sandstone of Brown Clee Hill and the adjacent area. 2. Palaeontology. Bulletin of the
British Museum (Natural History), Geology Series, 5, 243-310.

1985. The Cephalaspids from the Dittonian section at Cwm Mill, near Abergavenny, Gwent. Bulletin of
the British Museum (Natural History), Geology Series, 37, 149-171.
young, G. c. 1981. Biogeography of Devonian vertebrates. Alcheringa, 15, 225-243.

D. L. DINELEY
Department of Geology

Typescript received 6 September 1992 University of Bristol

Revised typescript received 19 May 1993 Bristol BS8 1RJ, UK

THE OSTRACOD GENUS KRITHE FROM THE
TERTIARY AND QUATERNARY OF THE
NORTH ATLANTIC

by G. P. COLES, R. C. WHATLEY and A. MOGUILEVSKY

Abstract. The ostracod genus Krithe is investigated from Cainozoic deep water sediments of the North
Atlantic region, mostly from ODP/DSDP cores, although other material is from short cores collected by
various NATO vessels, and the geographical and stratigraphical occurrence of the principal species is detailed.
The morphological features of the genus are described, and their relative utility in specific discrimination
assessed. Of the seventeen species and subspecies described herein, eight have been previously described, five
species ( Krithe gobanensis , K. regulare, K. minima, K. aquilonia, K. praemorkhoveni) and two subspecies
( K . morkhoveni lamellalata , K. morkhoveni ayressi) are new, while two are compared to previously described
species. In addition, the stratigraphical ranges and distribution of nineteen nomina nuda species, too rare to be
formally described, are tabulated. A new system, based on the nature of the anterior radial pore canals, is used
to erect distinct species groups of the genus and as an aid in species determination. The various carapace
biocharacters of the genus are reviewed and evaluated with respect to their taxonomic significance; many are
shown to be of little use since they are very conservative within Krithe while others, such as size, shape, degree
of dimorphism and the nature of the anterior radial pore canals are shown to be of great importance. The
stratigraphical ranges of all known Krithe species in the North Atlantic are given and the considerable
biostratigraphical significance of the genus is demonstrated. The previous use of Krithe in palaeoenvironmental
reconstruction is discussed and the supposed relationship between vestibulum morphology and oxygen level
critically evaluated and shown to be wanting, as is the proposition that there is a determinable relationship
between the length of species of the genus and water depth.

The ostracod genus Krithe Brady et al. (1874) is an abundant, cosmopolitan taxon with a
stratigraphical record of at least 97-5 m.y., extending from the Cenomanian (Shakin 1991) to the
Recent. Its diversity over this interval is indicated by the fact that at least 115 species have been
described (Kempf 1988). Many new species await description, especially from the deep sea, where
it reaches its maximum importance. It has been reported from all studies of bathyal and abyssal
faunas, since the voyage of H.M.S. Challenger (Brady, 1880) to the present investigation. In most
such studies, Krithe is both the most numerically abundant genus (sometimes outnumbering all
other ostracod taxa combined) and the most specifically diverse genus, where the whole ostracod
fauna is considered. Krithe also occurs in shallower waters on the continental shelf, where it is
dominantly, but not exclusively, cryophilic and is comparatively rare in subtropical waters. Some
unusual occurrences in shallow, warmer water environments where the genus is a dominant are
known, as in the Middle Miocene Balcomian of SE Australia (Whatley and Downing 1983).

However, the use of Krithe is either biostratigraphy or palaeoenvironmental analysis, capitalizing
upon its cosmopolitan and eurybathyal distribution, has largely been frustrated by the confused
state of its taxonomy. This is a direct result of the smooth carapace of all Krithe species and the
consequent lack of diagnostic ornamentation, together with an apparently high degree of
intraspecific variation, which together give rise to the so called ‘ Krithe Problem’ (McKenzie et al.
1990).

Krithe , and to a lesser extent Parakrithe van den Bold, have been used extensively in
palaeoenvironmental reconstructions, especially in deep water marine settings. This utilization has
been made based upon many claims concerning the biology and ecology of Krithe, the most

IPalaeontology, Vol. 37, Part 1, 1994, pp. 71-120, 7 pls.|

© The Palaeontological Association

72

PALAEONTOLOGY, VOLUME 37

important of which state that there exists an overall relationship between the size (length) of species
and water depth and also between the size and shape of the anterior vestibulum and levels of
dissolved oxygen. These hypotheses were formulated by Peypouquet (1975, 1977, 1979) and have
been considered in some detail and systematically rejected by Whatley and Zhao (1993). The present
study provides further evidence of the invalidity of these hypotheses.

Our taxonomic concept of Krithe is at considerable variance with that of Peypouquet (1975, 1977,
1979); using the techniques outlined below, we are able to recognize distinct species, using
essentially the same taxonomic database as Peypouquet. He recognized many fewer species, most
of which comprise a number of ecotypes, some of which varied in size considerably beyond what
we would consider possible within contemporary numbers of the same species, notwithstanding
differences in their ecology. As a general statement, it would seem that our species would often be
equivalent to Peypouquet’s ecotypes.

The present authors have studied Krithe in detail, many from Cenozoic deep-sea sediments in the
North Atlantic, in order to achieve a consistent taxonomy at the species level and to assess the
palaeoecological and biostratigraphical utility of the genus as a whole. We also employ the data of
a number of other researchers at the University of Wales, Aberystwyth in proposing a morphological
scheme for the identification of Krithe species.

MORPHOLOGICAL CHARACTERISTICS OF KRITHE

The external and internal morphological features of typical adult Krithe species are illustrated in
Text-figures 1-2. The smooth carapace of all Krithe species, and the consequent absence of much of
the information employed in the identification of ornate taxa, necessitates more detailed studies of
valve shape and size, as well as internal features only visible in well-preserved material. Nine
principal morphological features are considered, with particular emphasis upon their value in
specific discrimination.

Valve shape

Within certain parameters, this is quite constant within species, and is very useful in specific
discrimination once sexual dimorphism is recognized (Text-fig. 1). Males are relatively longer and
less high than females, but dimorphism, while often very marked, may be only weakly displayed in
some species. The general shape of Krithe varies from subovate through subquadrate to
subrectangular. The anterior margin is typically broadly rounded or slightly upswept, the posterior
margin is truncate, the dorsum is straight to convex and the ventral margin generally has a shallow
oral incurvature. However, there are many variations which complicate the basic generic outline.
For example, the posterior margin is especially variable; the postero-dorsal corner and posterior
angle (if present) vary considerably in their development. The postero-ventral slope makes an angle
of 60°-90° with the ventral margin, the stope being less steep in species with a highly tapered
posterior margin. The posterior indentation of the selvage is strongest in those valves with a tapered
posterior, and is inversely proportional in strength to the increasing angle of the postero-ventral
slope.

Valve size

This is often a useful specific character, as there is considerable size variation between species. The
smallest species recorded, Krithe sp. cf. K. parvula , is only 043 to 0 46 mm in length, while some
male specimens of K. trinidadensis exceed 1-50 mm. However, size is also influenced by sexual
dimorphism and intraspecific variation. Male valves are usually absolutely and always relatively
longer than females. Due to the strong overlap of most species of Krithe , one valve (usually the left)
is larger than the other.

The degree of intraspecific size variation for each species, based on the maximum dimensions of
both valves and both sexes, may be described as being small (< 10 per cent), moderate (10-20
per cent), considerable (20-30 per cent) to extremely large (over 30 per cent). Generally the largest

COLES ET AL.\ OSTRACOD GENUS KR1THE

73

Antero - dorsal concavity

Locellae Accommodation groove

Posterior indentation K .morkhoveni morkhoveni

of selvage [MLV]

text-fig. 1. Main morphological characteristics of the ostracod Krithe. a, d, K. pernoides sinuosus Ciampo,
male right valve, b-c, K. morkhoveni morkhoveni van den Bold, b, female right valve, c, male left valve.

species are the most variable in size (up to 50 per cent in the female valves of K. trinidadensis). In
contrast, the degree of size variation allowed by Peypouquet (1977) between contemporary
‘ecotypes’ of some of the species he recognized was considerably larger and reached a maximum
of 168 per cent within his Krithe sp. C. This degree of intraspecific size variation is quite unknown
among other Ostracoda and greatly exceeds that allowed by other workers.

Normal pore canals ( NPC)

These comprise approximately fourteen to eighteen round canal openings regularly distributed on
the surface of each valve. They are very constant in position in all Krithe species. They appear to
be a conservative character of little use in species discrimination. However, the normal pore canals
increase in size in proportion to valve size and, in the larger species, some are of sieve-type.

Inner lamella

The width of the inner lamella and the course of the line of concrescence vary considerably, both
between and within species. Species with the widest inner lamellae also have the most irregular lines
of concrescence, indented at the base of the radial pore canals and deeply depressed below the
adductor muscle scars. In species with narrow inner lamellae, the line of concrescence is usually
simple and is only slightly depressed below the adductors. The nature of the inner lamella is
generally a useful character in species identification.

Anterior vestibulum

The size and shape of the anterior vestibulum is a very useful specific character, notwithstanding
that it is moderately variable within certain species. The anterior vestibulum may be very large,
comprising most of the anterior marginal area, or it may be almost absent in some species. The
shape of the anterior vestibulum is more constant than its absolute size, and can be described in
terms of width of opening, length and width of neck and distal expansion. Four main types of

74

PALAEONTOLOGY, VOLUME 37

A

TYPE 1 A

K .aequabilis [MRV]

TYPE 2B

K .dolichodeira [MRV]

TYPE 3 A

K .pernoides sinuosa [MRV]

TYPE IB
K.reversa [MRV]

TYPE 2C
K. minima [MRV]

TYPE 3B

Krithe sp.13 [ARV]

K .morkhoveni lamellalata ’Y’ - shaped
[FRV] k .trinidadensis IFRV]

‘Pocket’- shaped [small]

K .pernoides sinuosa [MRV]

‘Pocket’ - shaped [large]
Krithe sp. 17 [ARV]

’Mushroom' - shaped

Krithe sp. [ARV]

‘Crescent’ - shaped
K. minim a [FRV]

text-fig. 2. Morphological characteristics of the anterior end of the Krithe valves, a, main types of antero-
dorsal radial pore canals (ADRPC). b, main types of anterior vestibule.

anterior vestibule can be distinguished (see Text-fig. 2b). Several intermediate examples occur, but
most specimens can be readily assigned to one of the four main types. In general, species with wide
inner lamellae have small, constricted anterior vestibulae, while species with narrow inner lamellae
have large, expanded anterior vestibulae.

Posterior vestibulum

This is usually much smaller than the anterior vestibulum and is subrectangular in shape. The
posterior vestibulum is relatively constant in shape but, since it varies little between species and is
frequently obscured by sediment infill, it is not useful as a specific diagnostic character.

Radial pore canals

The number, length and distribution of radial pore canals (both true and false) are very useful in
specific discrimination. The length of the radial pore canals (RPC) is directly proportional to the
width of the inner lamella. The RPC pattern within a species is remarkably stable, and can be
conveniently subdivided into five fields.

1. Antero-dorsal RPC. These comprise a maximum of five false canals emergent from the
vestibulum deck, dorsal of the first true anterior RPC (A RPC) or in the antero-dorsal fused zone,
and can be numbered AD 1-5. AD 1 is short or absent, AD 2 and AD 3 may occur as normal pore

COLES ET AL.\ OSTRACOD GENUS KRITHE

75

canals or may be short or moderately elongate, AD 4 may be absent, short or is frequently the most
elongate, and AD 5 is present only where AD 4 is elongate, when it is about half the length of AD 4.
Three basic types of antero-dorsal RPC can be recognized.

Type 1 :

AD

1A:

AD

IB:

AD

Type 2:

AD

2A :

AD

2B :

AD

2C :

AD

Type 3:

AD

3A :

AD

3B:

AD

The principal

-3 very short or absent, AD 4 short and emergent from a wide vestibular neck

Atlantic are illustrated in Text-fig. 2a and have proved more useful in species recognition than the
other three types of RPC.

2. Anterior RPC. These comprise nine to fourteen true RPC, emergent from the anterior
vestibulum. Their length varies in inverse proportion to the size of the anterior vestibulum. Up to
eight short false RPC are interspersed between them.

3. Ventral RPC. These comprise seven to fifteen false canals, some of which may branch. They are
prominent in species with wide ventral inner lamellae.

4. Postero-dorsal RPC. These comprise four false canals.

5. Posterior RPC. These comprise up to eight short, false canals.

Hinge

The hinge of Krithe is essentially adont; the narrow dorsal margin of the smaller right valve is
accommodated in a smooth groove in the larger left valve (Text-fig. 1 c). It may be straight or
arched, depending on the convexity of the dorsal margin. In species with a strongly convex dorsum
the hinge groove partly overhung by the dorsal margin. In many species the right valve hinge bar
is raised and finely denticulate at the posterior end. There are up to ten minute denticles and
complementary locellae in the left valve. The hinge teeth are most prominent in the male valves of
large species, producing a pseudodont dentition. Opposite hinge arrangements occur in species with
reversed valve overlap, where the right valve is larger than the left. Reversed overlap species are rare
and are confined almost exclusively to water depths of 1000 m or more.

The hinge is constant in form within a species, but is only of slight taxonomic use, since the same
basic pattern recurs in many different species. The pseudodont dentition of many deep-sea Krithe
species is almost exactly similar to that of the shallow water, often phytal genus Parakrithella Hanai,
despite the fact that the pseudodont dentition of the latter genus was used by Hanai to discriminate
it from Krithe. The similarity of hingement suggests a close relationship between the two genera,
notwithstanding their present-day ecological separation.

Muscle scars

The typical central muscle pattern is a straight to slightly arcuate vertical row of four adductors,
increasing in size dorsally. The topmost scar is usually reniform or slightly dorsally indented, the
second and third scars are smaller, flatter and often biconcave, and the fourth is always smallest and

76

PALAEONTOLOGY, VOLUME 37

subovate. The frontal scar is usually trilobate, but rarely may be quadrilobate or divided into two
scars. The dorsal scars are usually difficult to discern. The size of the muscle scars is generally in
direct proportion to valve size, although K. dolichodeira has relatively small scars. The pattern of
muscle scars is almost as variable within a species as between different species, and hence is of only
slight taxonomic value.

In summary, therefore, among the main morphological characteristics of Krithe species, valve
shape and size, the anterior vestibulum and the antero-dorsal radial pore canals are very useful in
specific discrimination ; the normal pore canals, the posterior radial pore canals and the muscle scars
are of very little use; and the remaining characters are of intermediate value.

MATERIAL AND METHODS

The principal source of samples for this study is material from ten Deep Sea Drilling Project sites
in the North Atlantic from Legs 80, 82 and 94. The studied sites, locations, present day water depths
and stratigraphical intervals are detailed in Table 1. This is the same material employed in the
studies of Whatley and Coles (1987, 1990) and Coles and Whatley (1989) in which emphasis was
placed on ostracod taxa other than Krithe.

table I The location, present-day water depth (PDWD) and stratigraphical intervals of the studied DSDP
sites.

Site

Latitude

Longitude

PDWD

(m)

Stratigraphical interval

549

49° 05-28' N

13° 05-88' W

2515

U. Palaeocene-U. Oligocene

550

48° 30 91 ' N

13° 26-37' W

4420

L. Palaeocene-L. Eocene

558

37° 46-20' N

37° 20-61' W

3754

L. Oligocene-U. Miocene

563

33° 38-50' N

43° 46-04' W

3786

M.-U. Miocene

606

37° 20-32' N

35° 29-99' W

3007

L. Pliocene-Quaternary

607

41° 00-07' N

32° 57-44' W

3427

U. Miocene-Quaternary

608

42° 50-21' N

23° 05-25' W

3526

U. Miocene-Quaternary

609

49° 52-67' N

24° 14-29' W

3884

U. Miocene-Quaternary

610

53° 13-30' N

18° 53-21' W

2417

U. Miocene-Quaternary

611

52° 50-47' N

30° 19-58' W

3201

U. Miocene-Quaternary

A total of 306 samples from twenty-one holes drilled at the above ten sites were analysed and their
entire ostracod fauna picked. Over 20000 specimens of Krithe were recovered, of which almost
2800 were adult specimens with fully developed internal features.

The internal features of Krithe valves were drawn using the Projectina microscope at a standard
magnification of x 100. The valves were first cleaned using a brush or fine needle, then placed in
a cavity slide and immersed in water. The image was projected onto tracing paper and drawn.

SYSTEMATIC PALAEONTOLOGY

The following section details the taxonomy, morphology, and biostratigraphical and geographical occurrences
of all important Krithe species present in the Cainozoic of the North Atlantic Ocean, with notes on their
distribution in other regions. It is intended to serve as an example of the way in which the morphological
characters outlined above can be used to separate species of this genus.

Type material of all new species is housed in the Palaeontology Collections of the Natural History Museum,
London, to which the catalogue numbers prefixed OS refer. Catalogue numbers prefixed GC/NA/ refer to
specimens in the Coles Collection, Micropalaeontology Museum, University of Wales, Aberystwyth. In the

COLES ET AL.\ OSTRACOD GENUS KRITHE

77

text, only the numbers of adult specimens are given; there are usually many more juvenile specimens in the
collections. The following conventions are employed: F, female; M, male; LV, left valve; RV, right valve;
ARPC, anterior radial pore canals; ADRPC, antero-dorsal radial pore canals; PRPC, posterior radial pore
canals; NPC, normal pore canals; PDWD, present-day water depth.

Size of species is given as small, medium, large, etc. according to the scale given by Whatley (1970, p. 301).
Dimension are given as L, length; and H, height.

Stratigraphical ranges are given by system or subdivision of system and by standard nannofossil zones (NP,
Palaeogene; NN, Neogene).

Under the heading Materia l and distribution the number is given of adult specimens of each species which
the authors had available for study and the geographical and stratigraphical distribution of the species within
the various cores on which the study is based. Additional distributional information, as well as taxonomic
comment, is given under Remarks with respect to each species.

Class ostracoda Latreille, 1806
Order podocopida Muller, 1894
Suborder podocopina Sars, 1866
Superfamily cytheracea Baird, 1850
Family krithidae Mandelstam in Bubikyan, 1958
Genus krithe Brady et ah, 1874

Discussion. The species and subspecies of Krithe discussed below are arranged according to their
ADRPC type.

ADRPC TYPE IB

Krithe reversa van den Bold, 1958
Plate 1, figures 1-6; Text-figure 3a-d

1958 Krithe reversa van den Bold, p. 404, pi. 1, figs 4 a-g.

1959 Krithe sawanensis Hanai, p. 302, figs, 3-4; pi. 18, figs 3-7.

1977 Krithe sp. C Peypouquet [pars], p. 109, fig. 36 sp. C 34U ( forme inverse only).

1981 Krithe reversa van den Bold; van den Bold, p. 69, pi. 1, fig. 13 a-d.

1981 Krithe reversa van den Bold; Steineck, p. 362, pi. 2, fig. 12.

1983 Krithe sp. C30 Peypouquet; Benson and Peypouquet, p. 818, pi. 5, figs 4-5.

1990 Krithe sp. 4 Dingle et a/., p. 282, figs 17d, 18f, 22e.

Material and distribution. Thirty seven adult valves. Holes 549A (U. Miocene), Hole 558 (M.-U. Miocene),
Hole 563 (M.-U. Miocene), Hole 606 (U. Pliocene-Quaternary), Hole 606A (L. Pliocene-Quaternary),
Hole 607 (L. Pliocene-Quaternary), Hole 608 (L.-U. Pliocene), Hole 609B (U. Pliocene-Quaternary),
Hole 610 (U. Pliocene), Hole 61 ID (L. Pliocene).

Dimensions (mm).

L

H

FLV GC/NA/1 18 606A-14 L. Pliocene

104

049

FLV GC/NA/1 38 606-5 Quaternary

0-98

047

FRV GC/NA/51 606A-14 L. Pliocene

103

046

MLV GC/NA/102 606A-14 L. Pliocene

0-94

044

MRV GC/NA/44 606A-14 L. Pliocene

0-91

040

FLV GC/NA/39 606A-19 L. Pliocene

0-95

0-50

FLV GC/N A/101 606A-14 L. Pliocene

0-95

0-52

FRV GC/NA/188 606-16 L. Pliocene

0-97

0-56

MLV GC/NA/1 17 606AM Quaternary

M2

048

MRV GC/NA/131 606-11 L. Pliocene

109

0-52

Stratigraphical range. Middle Miocene-Quaternary (NN 6-9, 14-16, 18-21).

Diagnosis. A large to very large, thick-shelled, subrectangular species of Krithe characterized by
reversed valve overlap and strong sexual dimorphism. Female much shorter and relatively higher

78

PALAEONTOLOGY, VOLUME 37

than the elongate male. Inner lamella quite narrow. Anterior vestibulum very large, ‘ pocket ’-
shaped, with wide opening, often extending to antero-dorsal corner of inner margin. Ten or eleven
straight to slightly curved ARPC with up to five false ARPC. ADRPC type IB. RV hinge bar
denticulate at posterior end.

Remarks. Both adults and juveniles of K. reversa are readily identifiable by their reversed valve
overlap and characteristic outline. The species occurs consistently from the Middle Miocene to the
Recent, but is never the dominant Kritlie species in any assemblage. The present is the earliest record
of this species in the North Atlantic. It is also present in the late Quaternary of the Northeast
Atlantic between PDWD 3422 and 4096 m (Porter 1984; material seen by GC), late Quaternary of
the Iberian Portal region at PDWD 3700 m (Harpur 1985), Middle Miocene to Pliocene (NN 9-18)
of numerous Caribbean localities in Trinidad, Haiti, Venezuela, Jamaica and the Dominican
Republic (van den Bold 1958, 1981; Steineck 1981), late Pliocene of Japan (Hanai 1959), latest
Pliocene and Quaternary of the Southwest Pacific and eastern Indian Ocean (Downing 1985; Ayress
1988) and the Upper Miocene to Quaternary of the Rio Grande Rise in the South Atlantic (Benson
and Peypouquet 1983).

In the Recent it is present in the Atlantic oft' Florida between 803 and 1080 m (among Cronin’s
1983 material seen by GC), the Gulf of Mexico below 1000 m (van Morkhoven, 1972; van den Bold,
1981), South Scotia Sea between 1 155 and 3925 m (Coxhill 1985), South Atlantic off Southwestern
Africa between 1600 and 2916 m (Dingle et al. 1990) and the North Atlantic from 803 to 5726 m
(personal observations).

In summary, K. reversa appears to have been present throughout the world’s oceans from at least
the Middle Miocene to the Recent and is usually indicative of water depths in excess of 1000 m. It
is absent from the Mediterranean Basin and southern Europe, suggesting that the water depths at
the Strait of Gibraltar were never sufficiently deep to permit its migration from the Atlantic.

ADRPC TYPE 2B
Krithe aequabilis Ciampo, 1986
Plate 1, figures 7-12; Text-figure 3e-k

1980 Krithe sp. 3 Ciampo, p. 17, pi. 4, fig. 4 (male).

1986 Krithe aequabilis Ciampo, p. 87, pi. 17, figs 1-2.

EXPLANATION OF PLATE 1

Figs 1-6. Krithe reversa (van den Bold). 1, GC/NA/39; DSDP Site 606A, c.c. 19; Early Pliocene; female left
valve, external view; x 44. 2-3, GC/NA/188; DSDP Site 607, c.c. 16; Early Pliocene; female right valve,
external and internal views; x43. 4-5, GC/NA/117; DSDP Site 606A, c.c. 4; Quaternary; male left valve,
external and internal views; x 37. 6, GC/NA/131 ; DSDP Site 606, c.c. 1 1 ; Late Pliocene; male right valve,
external view; x 39.

Figs 7-12. Krithe aequabilis Ciampo. 7-8, GC/NA/190; DSDP Site 607, c.c. 18; Early Pliocene; female left
valve, external and internal views; x47. 9, GC/NA/191; DSDP Site 607, c.c. 18; Early Pliocene; female
right valve, external view; x48. 10-11, GC/NA/49; DSDP Site 606A, c.c. 15; Early Pliocene; male right
valve, external and internal views; x 50. 12, GC/NA/172; DSDP Site 607, c.c. 10; Late Pliocene; male left
valve, internal view; x46.

Figs 13-18. Krithe dolichodeira van den Bold. 13-14, GC/NA/78; DSDP Site 606A, c.c. 13; Late Pliocene;
female left valve, external and internal views; x 53. 15, GC/NA/65; DSDP Site 606A, c.c. 9; Late Pliocene;
female right valve, external view; x 55. 16, GC/NA/16; DSDP Site 606A, c.c. 3; Quaternary; male left
valve, external view; x47. 17-18, GC/NA/45; DSDP Site 606A, c.c. 3; Quaternary; male right valve,
external and internal views; x47.

All figures are scanning electron micrographs.

PLATE 1

COLES et al., Krithe

80

PALAEONTOLOGY, VOLUME 37

Material and distribution. One hundred and sixty eight adult valves. Hole 549A (M.-U. Eocene), Hole
549A (U. Eocene-U. Oligocene), Hole 558 (L. Oligocene-L. Miocene), Hole 563 (M.-U. Miocene),
Hole 606 (Quaternary), Holes 606A and 607 (Pliocene), Hole 608 (Pliocene-Quaternary), Hole 609 B
(U. Pliocene-Quaternary), Holes 610 and 610C (Quaternary), Hole 610B (U. Pliocene).

Stratigraphical range. Middle Eocene to Quaternary (NP 16-25; NN 3, 5, 9, 14—21).

Diagnosis. A medium to large, sexually dimorphic species of Krithe, with ADRPC type 2B.
Moderately wide inner lamella with numerous RPC, eleven to twelve long ARPC, anterior
vestibulum medium-sized with scalloped margin. Males relatively and absolutely longer than
contemporaneous females.

Remarks. Ciampo (1986) described K. aequabilis from the late Miocene (Tortonian and Messinian)
of Italy. This species, recorded under several MS names, is abundant and widespread in the world’s
oceans. It resembles K. dolichodeira van den Bold, 1946, but differs in the expansion of the inner
lamella, producing a wider marginal zone. This results in a smaller anterior vestibulum with longer
ARPC, and modifies the ADRPC pattern so that the NPC between AD 1 and AD 3 appears as a
short AD 2.

Thus, K. aequabilis has ADRPC type 1A and K. dolichodeira has ADRPC type 2B. However,
AD 2 may be absent in some specimens, which are assigned to K. aequabilis on the basis of their
outline and anterior vestibule form. K. aequabilis generally has a more regularly curved posterior
than K. dolichodeira , although intermediate specimens exist. K. aequabilis occurs as two size morphs
in the Upper Eocene to Upper Oligocene of Site 549; the larger forms are 10-16% longer than the
smaller forms in the same samples. However, the two forms are not treated as distinct subspecies
as a near continuous series of intermediate specimens exist. K. aequabilis has been recorded from
several localities detailed below, but the total distribution is probably under-represented as it may
be included within K. dolichodeira by other authors. It is also recorded in MS only from the Middle
Miocene to Quaternary of the SW Pacific (Smith 1983; Dainty 1984; Downing 1985; Ayress 1988)
and Quaternary of the eastern Indian Ocean (Ayress 1988).

This species is also known from the Lower Oligocene of the Bay of Biscay (among the material
of Ducasse and Peypouquet 1979, seen by GC), Messinian of Sicily (Ciampo 1980), late Quaternary
of the Iberian Portal between PDWD 1200 and 3700 m (Harpur 1985), late Quaternary of the
Northeastern Atlantic in cores L4, R3 and R4 between latitude 43° and 61° North and PDWD 2177
and 3422 m (among the material of Porter 1984 seen by GC), Pleistocene and Recent of the Iberian

Dimensions (mm).

L H

0-66 0-31

0-63 0-30

0 71 0-34

0-62 0-29

0-89 0-45

0-88 0-44

0-66 0-30

0-64 0-28

0-73 0-34

0 71 0-31

0-87 0-42

0-76 0 31

0-68 0-29

0-92 0-39

0-78 0-30

0-75 0-30

0-76 0 41

0-67 0-26

0-85 0-35

FLV GC/NA/863 549A 34-1 U. Eocene
FLV GC/NA/864 549A 6-5 U. Oligocene
FLV GC/NA/865 549A 8-5 U. Oligocene
FLV GC/NA/876 549A 8-5 U. Oligocene
FLV GC/NA/190 607-18 L. Pliocene
FLV GC/NA/71 606A-9 U. Pliocene
FRV GC/NA/866 549 6^t M. Eocene
FRV GC/NA/867 549A 6-5 U. Oligocene
FRV GC/NA/868 549A 15-1 L. Oligocene
FRV GC/NA/869 549A 9-5 U. Oligocene
FRV GC/NA/191 607-18 L. Pliocene
MLV GC/NA/870 549A 17-1 U. Eocene
MLV GC/NA/871 549A 6-5 U. Oligocene
MLV GC/NA/172 607-10 U. Pliocene
MRV GC/NA/872 549A 9-5 U. Oligocene
MRV GC/NA/873 549A 17-1 U. Eocene
MRV GC/NA/874 549A 1 1-5 L. Oligocene
MRV GC/NA/875 549A 6-5 U. Oligocene
MRV GC/NA/49 606A-15 L. Pliocene

COLES ET AL.: OSTRACOD GENUS KRITHE

Portal between PDWD 1285 and 2798 m (Elant 1985), Recent North Atlantic off Mauritania
between 3856 and 4000 m (Barkham 1985), Upper Miocene of San Marino and Majorca and
Quaternary of the Hebrides Terrace Seamount in the eastern North Atlantic at PDWD 1250 m
(personal observations).

In summary, K. ciequabilis is common in the Middle Eocene to Recent of the North Atlantic,
although it is usually less abundant than K. dolichodeira and has a narrower depth range (1200 to
4000 m compared with 200 to 5440 m).

Krithe dolichodeira van den Bold, 1946

Plate 1, figures 13-18; Text-figure 3l-q

71880 Ilyobates compressa Seguenza, p. 325, pi. 17, figs 30, 30a (male).

71880 Krithe producta Brady [pars], p. 114, pi. 27, fig. i only (female).

1946 Krithe dolichodeira van den Bold, p. 75, pi. 4, fig. 14 a-b (male).

1946 Krithe praetexta (Sars); van den Bold, p. 75, pi. 4, fig. 16 a-b (female).

1964 Krithe cf. bartonensis (Jones); Dieci and Russo, p. 78, pi. 15, fig. 15 a-b (female).

1964 Krithe interrupta Dieci and Russo, p. 92, pi. 2, fig. 5 a-b; pi. 4, fig. 5 a-b (males).

1964 Krithe monosteracensis (Seguenza); Ascoli, pi. 4, fig. 4 (female).

1967 Krithe oertlii Dieci and Russo, p. 15, pi. 3, figs 7-8 (male).

1967 Krithe bartonensis (Jones); Hidings, p. 644, fig. 5 a (male).

1968 Krithe dolichodeira van den Bold; van den Bold, p. 51, pi. 2, figs 9 a-b, 12 a-b\ pi. 10 , fig. 4a-d.

1968 Krithe dolichodeira van den Bold; Russo, p. 37, pi. 6, figs 2 a-b, 2dl (male).

1969 Krithe producta Brady; Yassini, p. 82, pi. 22, fig. 4 a (male).

1972 Krithe monosteracensis (Seguenza); Sissingh, p. 84, pi. 4, fig. la-b (female).

1975 Krithe monosteracensis (Seguenza); Breman, p. 209, pi. 1, fig. la (male), 1 b (female).

1976 Krithe monosteracensis (Seguenza); Breman, p. 54, pi. 3, fig. 29.

1977 Krithe sp. C Peypouquet [pars], p. 109, sp. C13fn, C22fn, C23fn, C24fn, C34en (all female).

1981 Krithe dolichodeira van den Bold; van den Bold, p. 66, pi. 1, fig. 12 (male).

1981 Krithe vandenboldi Steineck, p. 358, fig. 12, pi. 1, figs 1-4 (male).

1983 Krithe sp. C22 Peypouquet; Benson and Peypouquet, p. 818, pi. 5, fig. 2.

1984 Krithe vandenboldi Steineck; Steineck et al., p. 1473, fig. 9 j-k (males).

1990 Krithe spatularis Dingle et al., p. 272, figs I6d-f, 17b, 18e (males).

Material and distribution. Three hundred and eighty adult valves. Hole 549 (L.-U. Eocene), Hole 549A
(U. Eocene-U. Oligocene, U. Miocene), Hole 558 (L. Oligocene-U. Miocene), Hole 563 (M. Miocene-U.
Miocene), Holes 606, 606A and 607 (Pliocene-Quaternary), Holes 608 and 608A (U. Pliocene-Quaternary),
Holes 609 and 609B (L. Pliocene, Quaternary), Hole 610 (Miocene-Quaternary), Hole 610B (U. Pliocene),
Hole 610C (Quaternary), Hole 61 ID (L. Pliocene).

Dimensions {mm).

L

H

FLV GC/NA/853 549 15-1 L. Eocene

0-60

0-29

FLV GC/N A/854 549A 16-1 L. Oligocene

0-82

0-39

FLV GC/NA/78 606A-13 U. Pliocene

0-80

0-42

FLV GC/N A/74 606A-6 Quaternary

0-81

0-40

FRV GC/NA/855 549 15-5 L. Eocene

0-63

0-30

FRV GC/NA/856 558 14-5 L. Miocene

0-71

0-32

FRV GC/NA/857 558 17-5 L. Miocene

0-73

0-33

FRV GC/N A/858 549 15-1 L. Eocene

0-60

0-28

FRV GC/NA/65 606A-9 U. Pliocene

0-76

0-38

MLV GC/NA/859 549 2-1 M. Eocene

0-70

0-30

MLV GC/N A/860 549 H3-3 U. Eocene

0-71

0 33

MLV GC/NA/861 563 10-1 M. Miocene

0-70

0-30

MLV GC/NA/16 606A-3 Quaternary

0-89

0-39

MRV GC/NA/862 549A 8-5 L. Oligocene

0-71

0-30

MRV GC/NA/45 606A-3 Quaternary

0-89

0-38

Stratigraphical range. Eocene-Quaternary (NP 10-19, 21 25; NN 1-9, 11-21).

PALAEONTOLOGY, VOLUME 37

text-fig. 3. For legend see opposite.

COLES ET A L.\ OSTRACOD GENUS KRITHE

83

Diagnosis. A medium to large species of Krithe with strongly pronounced sexual dimorphism. Males
subrectangular and much longer than females, which are more subquadrate and slightly more
inflated posteriorly. Both sexes with dorsal margin almost straight, parallel to ventral margin.
Posterior truncate with posterior concavity, strongest in male. Inner lamella narrow, inner margin
early parallel with outer margin. Large ‘mushroom ’-shaped anterior vestibule with wide opening,
short neck, frequently extending towards the antero-dorsal corner of the inner margin. ADRPC
type 2B; ten to eleven very short, regularly spaced ARPC in fan arrangement.

Remarks. This is an abundant, long-ranging species which shows some variation in size and anterior
vestibulum shape. There is not a simple increase in size with time, as Upper Eocene and Quaternary
specimens may be of similar size, and the largest specimens occur in the Lower Oligocene. Van den
Bold (1946) described this species as K. dolichodeira from the Miocene of Cuba, but gave poor
illustrations. Subsequent authors have, however, used the name K. dolichodeira for species clearly
conspecific with the present material (e.g. Russo 1968). Other authors have identified this species as
K. monosteracensis or K. compressa , both described by Seguenza (1880) from the Pliocene of Sicily.
The present material is not conspecific with K. monosteracensis as figured by Seguenza, but is similar
in shape to the figures of K. compressa. However, as Seguenza’s types were destroyed in an
earthquake, the identity of these species cannot be confirmed and, therefore, the name K. compressa
is not used here. K. dolichodeira differs from K. aequabilis Ciampo, 1986, in the size and shape of
the anterior vestibule, RPC pattern and inner lamella. K. interrupt a Died and Russo, 1964, is
included within K. dolichodeira since it seems to represent an extreme male variant of this species
with a very caudate posterior, and also because intermediate specimens occur.

Krithe producta Brady 1880 should be considered a nomen dubium. The lectotype of Puri and
Hulings (1976) is a juvenile valve of indeterminate species and the original illustrations of Brady
(1880) embrace some five species, including those with both normal and reversed overlap.

text-fig. 3. a-d. K. reversa (van den Bold). A, GC/NA/39; DSDP Site 606A, c.c. 19; Early Pliocene; female
left valve, b, GC/NA/188; DSDP Site 607, c.c. 16; Early Pliocene; female right valve, c, GC/NA/131 ; DSDP
Site 606, c.c. 1 1 ; Late Pliocene; male right valve, d, GC/NA/117; DSDP Site 606A, c.c. 4; Quaternary; male
left valve, e-k. K. aequabilis Ciampo. e, GC/NA/191 ; DSDP Site 607, c.c. 18; Early Pliocene; female right
valve, f, GC/NA/190; DSDP Site 607, c.c. 18; Early Pliocene; female left valve. G, GC/NA/865; DSDP Site
549A, c.c. 8-5; late Oligocene; female left valve, h, GC/NA/868; DSDP Site 549A, c.c. 15-1 ; early Oligocene;
female right valve, i, GC/NA/172; DSDP Site 607, c.c. 10; Late Pliocene; male left valve, j, GC/NA/49;
DSDP Site 606A, c.c. 15; Early Pliocene; male right valve, k, GC/NA/873; DSDP Site 549A, 17-1; late
Eocene; male right valve, l-q. K. dolichodeira van den Bold, l, GC/NA/853; DSDP Site 549, cc. 15-1 ; early
Eocene; female left valve, m, GC/NA/78; DSDP Site 606A, c.c. 13; Early Pliocene; female left valve.
n, GC/NA/65; DSDP Site 606A, c.c. 9; Late Pliocene; female right valve, o, GC/NA/16; DSDP Site 606A,
c.c. 3; Quaternary; male left valve, p, GC/NA/45; DSDP Site 606A, c.c. 3; Quaternary; male right valve.
Q, GC/NA/862; DSDP Site 549A, c.c. 8-5; late Oligocene; male right valve. R-v. K. gobanensis sp. nov. R,
OS 14015; DSDP Site 548, c.c. 7-4; middle Eocene; female left valve, s, OS 14022, DSDP Site 549A, c.c. 27 I ;
late Eocene; male right valve, t, OS 14020; DSDP Site 548, c.c. 7-1 ; middle Eocene; female right valve, u,
OS 14021, DSDP Site 549A, c.c. 1 1-2; early Oligocene; female right valve, v, OS 14016; DSDP Site 548A, c.c.
11-2; early Oligocene; female left valve, w-z. K. regulare sp. nov. w, OS 14002; DSDP Site 548 A, c.c. 8-5;
late Oligocene; female left valve, x, OS 14010; DSDP Site 549A, c.c. 8-5; late Oligocene; female right valve.
y, OS 14012; DSDP Site 549, c.c. 8-5; late Oligocene; male right valve, z, OS 14006; DSDP Site 549, c.c. H3-2;
late Eocene; male left valve, aa-dd. Krithe sp. cf. K. hiwanneensis Elowe and Law. aa, GC/NA/878; DSDP
Site 549A, c.c. 8-5; late Oligocene; female left valve, bb, GC/NA/882; DSDP Site 549A, c.c. 8-5; late
Oligocene; female right valve, cc, GC/NA/880; DSDP Site 549A, c.c. 17-1; late Eocene; male left valve.
dd, GC/NA/885; DSDP Site 549A, c.c. 8-5; late Oligocene; male right valve, ee-jj. K. minima sp. nov. ee,
OS 13954; DSDP Site 606A, c.c. 14; Early Pliocene; female left valve, ff, OS 13955; DSDP Site 606A, c.c. 14;
Early Pliocene; female right valve, gg, OS 13958; DSDP Site 606A, c.c. 17; Early Pliocene; male left valve.
hh, OS 13959; DSDP Site 606A, c.c. 17; Early Pliocene; male right valve, n, OS 13956; DSDP Site 606A,
c.c. 5; Quaternary; male left valve, jj, OS 13957; DSDP Site 606, c.c. 9; Late Pliocene; male right valve.

All projectina drawings; x 50.

84

PALAEONTOLOGY, VOLUME 37

This species is extremely abundant and widespread from the Eocene to the Recent in the Atlantic,
Caribbean and Mediterranean. It is tolerant of a wide depth range, extending from the outer shelf
at 200 m to the calcium compensation depth (CCD, maximum recorded depth of species 5440 m),
with most records from bathyal and abyssal depths. The following occurrences are noted : Eocene
of DSDP Sites 612 and 613 off New Jersey (material of Cronin and Compton-Gooding 1987; seen
by GC), Middle Miocene to Pliocene deep water sediments throughout the Caribbean (van den Bold
1977, 1981), Oligocene of Barbados (Steineck et al. 1984), Miocene and Pliocene of the Central
Equatorial Pacific (Steineck et al. 1988), Miocene of Jamaica (Steineck 1981), Middle Miocene to
Quaternary of the Rio Grande Rise, South Atlantic (Benson and Peypouquet 1983), Upper
Miocene of Italy (Dieci and Russo 1964, 1967; Russo 1968), Pliocene of Crete (Sissingh 1972),
Pleistocene of the Western Mediterranean and Iberian Portal region between PDWD 795 and
2798 m (Elant 1985), late Quaternary of the Iberian Portal region between PDWD 1200 and 3700 m
(Harpur 1985) and the late Quaternary of the eastern North Atlantic in cores L4, L5, 01, 02, P2,
R2, R3, R4, R5 and S3 between latitude 43° and 64°N and PDWD 938 and 4556 m (among the
material of Porter 1984 seen by GC).

Occurrences in Recent sediments comprise the Gulf of Mexico between 200 m and 800 m (van den
Bold 1981) and 1500 m (personal observations), throughout the North Atlantic between 210 and
5440 m (personal observations; Peypouquet 1977; Davies 1981; Barkhain 1985), slope off Florida
between 579 and 739 m (among the material of Cronin 1983 seen by GC). Bay of Biscay between
200 and 3900 m (Yassini 1969), Adriatic between 243 and 1200 m (Breman 1975, 1976) and the
western Mediterranean and Iberian Portal area from 585 and 2798 m (Elant 1985). Dingle et al.
(1990) describe this species as K. spatularis from the Recent off southwestern Africa between water
depths of 392 and 1662 m.

Krithe gobanensis sp. nov.

Plate 2, figures 1-3; Text-figure 3r-v

Derivation of name. From the only known occurrence of this species on the Goban Spur.

Holotype. Female left valve, OS 14015.

Type locality and horizon. DSDP Site 549, Hole 549, Goban Spur, lat. 49° 05 28' N ; long. 1 3° 05-88' W ; PDWD
2513 nr. Core 7, section 4, interval 0-88-0-95 m. Middle Eocene, NP 15. Pale greenish-white nannofossil ooze.

EXPLANATION OF PLATE 2

Figs 1-3. Krithe gobanensis sp. nov. 1-2, OS 14018; DSDP Site 549A, c.c. 1 1-2; early Oligocene; female left
valve, external and internal views; x 65. 3, OS 14019; DSDP Site S49A, c.c. 1 1-2; early Oligocene; female
right valve, external view; x65.

Figs 4-9. Krithe regularae sp. nov. 4-5, OS 14004; DSDP Site 549A, c.c. 2-5; late Oligocene; female left valve,
external and internal views; x 68. 6, OS 14008; DSDP Site 549A, c.c. 8-5; late Oligocene; female right valve,
external view; x68. 6, OS 14008; DSDP Site 549A, c.c. 8-5; late Oligocene; female right valve, external,
view; x 66. 7, OS 14007; DSDP Site 549A, c.c. 8-5; late Oligocene; male left valve, external view; x68.
8-9, OS 140011; DSDP Site 549A, c.c. 8-5; late Oligocene; male right valve, external and internal
views; x66.

Figs 10-15. Krithe sp. cf. K. hiwanneensis Howe and Law. 10, GC/NA/878; DSDP Site 549A, c.c. 8-5; late
Oligocene; female left valve, external view; x 78. 11-12, GC/NA/882; DSDP Site 549A, c.c. 8-5; late
Oligocene; female right valve, external and internal views; x75. 13-14, GC/NA/881; DSDP Site 549A,
c.c. 8-5 ; late Oligocene ; male left valve, external and internal views ; x 70. 1 5, GC/NA/885 ; DSDP Site 549A,
c.c. 8-5; late Oligocene; male right valve, external view; x68.

Figs 16-18. Krithe minima sp. nov. 16, OS 13962; DSDP Site 606A, c.c. 7; Late Pliocene; female left valve,
external view; x 66. 17 18, OS 13961 ; DSDP Site 606A, c.c. 5; Quaternary; female right valve, external and
internal views; x66.

All figures are scanning electron micrographs.

PLATE 2

COLES et al Krithe

86

PALAEONTOLOGY, VOLUME 37

Material and distribution. Ninety eight adult valves. Hole 549 (?U. Palaeocene, M.-U. Eocene), Hole 549A
(U. Eocene-U. Oligocene).

Dimensions {mm).

Holotype FLY OS 14015 549 7-4 M. Eocene
Paratype FLV OS 14016 549A 11-2 L. Oligocene
Paratype FLV OS 14017 549 19-3 U. Palaeocene
Paratype FLV OS 14018 549A 11-2 L. Oligocene
Paratype FRV OS 14019 549A 11-2 L. Oligocene
Paratype FRV OS 14020 549 7-1 M. Eocene
Paratype FRV OS 14021 549A 11-2 L. Oligocene
Paratype MRV OS 14022 549A 27-1 U. Eocene

L

H

0-56

0-30

0-64

0-35

0-57

0-32

0-64

0-34

0-64

0-32

0-57

0-30

0-62

0-32

0-66

0-28

Stratigraphical range. ? Upper Palaeocene, Middle Eocene to Upper Oligocene (NP 77, 15-23).

Diagnosis. A medium, subovate to elongate-subovate species of Krithe with a narrow inner lamella
and a larger anterior vestibulum which has a wide opening and is only slightly expanded distally.
ADRPC type 2B.

Description. Medium, subovate to elongate-subovate and moderately inflated. FLV dorsum slightly convex,
broadly rounded anterior, posterior bluntly truncate with steep postero-dorsal slope, ventral margin almost
straight. FRV as FLV but dorsum more convex with shallow antero-dorsal concavity. Males as females but
much more elongate with straighter dorsum. Normal LV > RV overlap. NPC small and widely scattered. Inner
lamella narrow, with simple line of concrescence. Anterior vestibulum large and ‘pocket ’-shaped with wide
opening and very slightly expanded distally. ADRPC type 2B, although a few specimens appear to have type
1A. Eleven simple ARPC in fan arrangement with a few short false canals. Hinge adont. Frontal scar may be
quadrifoil.

Remarks. This is a common species from the Middle Eocene to the base of the Upper Oligocene of
Site 549 to which it is apparently confined, with a single poorly preserved valve occurring in the
Upper Palaeocene. Almost all specimens are female; only three male valves were found. It resembles
K. regulare sp. nov., but that species is more elongate with an almost straight dorsum and a more
distally expanded anterior vestibulum.

Krithe regulare sp. nov.

Plate 2, figures 4-9; Text-figure 3w-z

Derivation of name. Latin, with reference to the regular, subrectangular outline of this species.

Holotype. Female left valve, OS 14002.

Material and distribution. One hundred and ninety three adult valves. Hole 549 (L.-U. Eocene), Hole 549A
(U. Eocene-U. Oligocene), Hole 558 (L. Oligocene).

Type locality and horizon. DSDP Site 549, Hole 549A, Goban Spur, lat. 49° 05-28' N; long. 13° 05-88' W,
PDWD 2513 m. Core 8, section 5, interval 0-80-0-87 m. Upper Oligocene, NP 24. Creamy white nannofossil
ooze.

Dimensions {mm).

L

H

Holotype FLV OS 14002 549A 8-5 U. Oligocene

0-62

0-30

Paratype FLV OS 14003 549 10-1 L. Eocene

0-64

0-30

Paratype FLV OS 14004 549A 8-5 U. Oligocene

0-62

0-31

Paratype MLV OS 14005 549A 8-5 U. Oligocene

0-62

0-28

Paratype MLV OS 14006 549 H3-2 U. Eocene

0-61

0-27

Paratype MLV OS 14007 549A 8-5 U. Oligocene

0-62

0-28

Paratype FRV OS 14008 549A 8-5 U. Oligocene

0-64

0-29

COLES ET AL.\ OSTRACOD GENUS KRITHE

87

Dimensions (mm).

Paratype FRV OS 14009 549A 13-2 L. Oligocene
Paratype FRV OS 14010 549A 8-5 U. Oligocene
Paratype MRV OS 14011 549A 8-5 L. Oligocene
Paratype MRV OS 14012 549 5-1 M. Eocene
Paratype MRV OS 14013 549 A 8-5 L. Oligocene
Paratype MRV OS 14014 549 A 9-5 L. Oligocene

Stratigraphical range. Lower Eocene to Upper Oligocene (NP 13-25).

L

H

0-65

0-31

0-64

0-29

0-64

0-26

0-67

0-27

0-63

0-25

0-60

0-25

Diagnosis. A medium, slightly sexually dimorphic, elongate-subrectangular species of Krithe with
almost parallel, straight dorsal and ventral margins. Posterior bluntly truncate with slight posterior
angle. Inner lamella narrow, anterior vestibulum large and ‘mushroom ’-shaped. ADRPC type 2B.

Description. Medium, elongate-subrectangular and slightly inflated carapace. FLV dorsum very slightly
convex, broadly rounded anterior, posterior bluntly truncate with steep postero-dorsal slope and no posterior
angle; ventral margin almost straight. FRV as FLV but dorsum slightly more convex. Male as female but more
elongate and with straighter dorsum. Normal valve overlap. Inner lamella narrow, line of concrescence simple
or with slight indentations at the base of the RPC. Anterior vestibulum large and ‘mushroom '-shaped with
wide opening and expanded distally with rounded margin. ADRPC type 2B. A few specimens have a very short
additional ADRPC arising from the base of AD 2. Twelve to fourteen quite short ARPC in fan arrangement,
very few false ARPC. Ventral RPC short and well-spaced. Posterior vestibulum large with very short PRPC.
Muscle scars relatively small; frontal scar may be subdivided.

Remarks. This species is common in the Middle Eocone to Upper Oligocene of Site 549, with two
specimens present in the upper Lower Eocene sample 549 10-1. It is most similar to K. gobanensis
and might be considered to be the male of that species. However, the two species differ in the shape
of their anterior vestibula, males and females can be consistently distinguished in K. regulare and
a few males of K. gobanensis are known. The two species also differ in stratigraphical range;
K. regulare is common in the Upper Oligocene samples 549A 9-5 to 549A 6-5 where K. gobanensis
is absent. K. regulare resembles K. dolichodeira in vestibular form and ADRPC pattern, but differs
in shape, having a more rounded posterior and much less marked sexual dimorphism.

Krithe sp. cf. K. hiwanneensis Howe and Law, 1936
Plate 2, figures 10-15; Text-figure 3aa-dd
71936 Krithe hiwanneensis Flowe and Law, p. 72, pi. 5, figs 32-4.

Material and distribution. Two hundred and sixteen adult valves. Hole 549 (L.-U. Eocene), Hole 549A
(U. Eocene-U. Oligocene), Hole 558 (L. Oligocene-L. Miocene), Hole 563 (M. Miocene).

Dimensions (mm).

FLV GC/NA/877 549 2-4 M. Eocene
FLV GC/NA/878 549A 8-5 U. Oligocene
FLV GC/NA/879 558 24-3 U. Oligocene
MLV GC/NA/880 549A 17-1 U. Eocene
MLV GC/NA/881 549A 8-5 U. Oligocene
FRV GC/NA/882 549A 8-5 U. Oligocene
FRV GC/NA/883 558 18-5 L. Miocene
MRV GC/NA/884 563 10-1 M. Miocene
MRV GC/NA/885 549A 8-5 U. Oligocene
MRV GC/NA/886 549 13-1 L. Eocene

L

H

0-57

0-33

0-54

0-30

0-71

0-41

0-59

0-29

0-60

0-29

0-56

0-29

0-54

0-28

0-60

0-28

0-62

0-28

0-58

0-28

Stratigraphical range. Lower Eocene to Middle Miocene (NP 11-25; NN 1, 5).

Diagnosis. A medium (one FLV is large), subrectangular species of Krithe with almost straight,
subparallel dorsal and ventral margins. Posterior bluntly truncate with shallow postero-dorsal

88

PALAEONTOLOGY, VOLUME 37

concavity and almost vertical postero-ventral slope. Inner lamella narrow, subparallel to outer
margin. Anterior vestibulum large, 'mushroom '-shaped with narrow neck and may extend up the
antero-dorsal inner margin. Eleven to thirteen ARPC short and evenly spaced, ADRPC type 2B.
Sexually dimorphic; males relatively and absolutely longer than contemporaneous females.

Remarks. This species is common from the Middle Eocene to Upper Oligocene of Site 549, but rare
elsewhere, particularly in the Miocene. One late Oligocene female specimen (GC/NA/879) is much
larger than typical specimens but is otherwise identical; it may represent a post-maturation moult.
The present specimens are very similar to the diagrammatic figures of K. hiwanneensis Howe and
Law, 1936, from the Oligocene of the North American Gulf Coast. However, they are only
compared to K. hiwanneensis as the length of the syntype was given as 0-70 mm, longer than all but
one of the present specimens, and because specimens of K. hiwanneensis illustrated by other authors
do not appear to be conspecific. K. sp. cf. K. hiwanneensis is similar to, and may have evolved from,
K. dolichodeira in the early Eocene. However, in any given sample, the present species is smaller than
K. dolichodeira , has a more bluntly truncate posterior, and a more constricted vestibular neck.

ADRPC TYPE 2C
Krithe minima sp. nov.

Plate 2, figures 16-18; Plate 3, figures 1-5; Text-figure 3ee-jj
71977 Krithe sp. Cl 1 fn Peypouquet, p. 109, fig. 36 [pars].

Derivation of name. Latin, with reference to the small size of this species.

Holotype. Female left valve, OS 13954.

Type locality and horizon. DSDP Site 606, Hole 606A, middle North Atlantic, southwest of the Azores,
lat. 37° 20-29' N; long. 35° 30 02' W. Core catcher 14, Lower Pliocene, NN 15. White nannofossil ooze.

Material and distribution. One hundred and three adult valves. Hole 558 (U. Miocene), Hole 563
(M.-U. Miocene), Holes 606 and 606A (Pliocene-Quaternary), Hole 607 (U. Miocene-Quaternary), Hole 608
(U. Miocene-Pliocene), Holes 608A and 610 (U. Pliocene-Quaternary), Holes 609B and 610C (Quaternary),
Hole 61 ID (L. Pliocene, Quaternary).

EXPLANATION OF PLATE 3

Figs 1-5. Krithe minima sp. nov. 1, OS 13962; DSDP Site 606A, c.c. 7; Late Pliocene; female left valve, internal
view; x 66. 2-3, OS 13956; DSDP Site 606A, c.c. 5; Quaternary; male left valve, external and internal views;
x 65. 4-5, OS 13957 ; DSDP Site 606, c.c. 9; Late Pliocene; male right valve, external and internal views; x 63.

Figs 6-10. Krithe crassicaudata van den Bold. 6-7, GC/NA/958; DSDP Site 549A, c.c. 39-2; late Eocene;
juvenile left valve, external and internal views; x 64. 8-9, GC/NA/959; DSDP Site 549A, c.c. 39-2; late
Eocene; juvenile right valve, external and internal views; x68. 10, GC/NA/956; DSDP Site 549A, c.c. 13-2;
early Oligocene; male right valve, internal view; x53.

Figs 11-18. Krithe morkhoveni morkhoveni van den Bold. 11, GC/NA/108; DSDP Site 606A, c.c. 14; Early
Pliocene; female left valve, external view; x 43. 12-13, GC/NA/1 14; DSDP Site 606, c.c. 15; Early Pliocene;
female right valve, external and internal views; x46. 14, GC/NA/121; DSDP Site 606A, c.c. 17; Early
Pliocene ; female left valve, internal view ; x 46. 15, GC /N A/ 118; DSDP Site 606A, c.c. 1 4 ; Early Pliocene ;
male left valve, external view; x40. 16-17, GC/NA/44; DSDP Site 606A, c.c. 14; Early Pliocene; male right
valve, external and internal views; x46. 18, GC/NA/102; DSDP Site 606A, c.c. 14; Early Pliocene; male
left valve, internal view; x45.

All figures are scanning electron micrographs.

PLATE 3

COLES et al. , Krithe

90

PALAEONTOLOGY, VOLUME 37

Dimensions (mm).

L

H

Holotype FLV OS 13954 606A-14 L. Pliocene

0-60

0-30

Paratype FLV OS 13955 563 10-1 M. Miocene

0-60

0-28

Paratype FLV OS 13956 606A-7 U. Pliocene

0-64

0 31

Paratype MLV OS 13957 606A-5 Quaternary

0-65

0 31

Paratype MLV OS 13958 606A-17 L. Pliocene

0-60

0-28

Paratype FRV OS 13959 606A-14 L. Pliocene

0-59

0-28

Paratype FRV OS 13960 606A-5 Quaternary

0-64

0-30

Paratype MRV OS 13961 606-9 U. Pliocene

0-67

0-28

Paratype MRV OS 13962 606A-17 L. Pliocene

0-59

0-25

Stratigraphical range. Middle Miocene-Quaternary (NN 5, 9, 11-12, 14-21).

Diagnosis. A small, elongate, sexually dimorphic species of Krithe with a small, subcrescentic
anterior vestibulum and ADRPC type 2C.

Description. Small, elongate subrectangular and moderately inflated. Sexually dimorphic; females relatively
shorter than males, with a slightly convex, rather than a straight dorsum. Anterior margin broadly rounded,
ventral margin with slight oral incurvature, posterior truncate with shallow posterior concavity and steep
postero-ventral slope. Thin-shelled, LV slightly overlaps RV. Inner lamella narrow; line of concrescence
slightly sinuous and subparallel with outer margin. NPC small and closely spaced. Anterior vestibulum small,
subcrescentic, upswept with narrow opening. Eleven short, straight ARPC in fan arrangement. ADRPC type
2C; posterior vestibulum relatively large, with up to five PRPC. Hinge pseudodont, RV hinge bar finely
denticulate at posterior end. Central muscle scars small, consisting of a slightly arcuate row of four adductors
with topmost scar dorsally indented and trefoil frontal scar.

Remarks. K. minima is rare in the Middle and Upper Miocene, but is more common in the Pliocene
and Quaternary, although it is never a dominant species. It is recorded (mainly in manuscript) from
the North Atlantic; the failure of other authors to note it may be due to its small size and inclusion
with juvenile specimens of other, larger species. It occurs in the Lower Pliocene of DSDP Site 406
on the Rockall Plateau (among the material of Ducasse and Peypouquet 1979 seen by GC); late
Quaternary of the Northeastern Atlantic between latitudes 43° and 68° N and PDWD between 1678
and 4566 m in cores L4, L5, N3, P2, R2, R3, R4, R5, S3, T2 and T3 (among the material of Porter
1984 seen by the authors); late Quaternary of the Iberian Portal between PDWD 1200 and 3700 m
(Harpur 1985); Quaternary of the western Mediterranean and Gulf of Cadiz between PDWD 900
and 2798 m (among the material of Elant 1985 seen by GC); Recent Atlantic off Florida at 739 and
472 m among the material of Cronin 1983 seen by GC), the Upper Miocene of San Marino and the
Quaternary of the Hebrides Terrace Seamount at 1250 m (personal observations). The single Pacific
record is that of Smith (1983) from the late Quaternary of the Challenger Plateau at a PDWD of
1066 m.

K. sp. Cll fn of Peypouquet (1977) appears to be the male of K. minima , although the length
range quoted (0-32-0-38 mm) is much smaller than any known Krithe species and is probably an
error. In summary, K. minima is widespread in the Upper Miocene to Recent of the North Atlantic
between 28° and 68° N, with most specimens recorded between 1000 and 3000 m. It is also present
in the Quaternary of the Mediterranean at bathyal depths. It is apparently unrecorded from the
Caribbean region, although this may be due to its relatively small size which renders it easily
overlooked among the juveniles of other species of the genus.

ADRPC TYPE 3A

(Group K. trinidadensis van den Bold, 1958)

Discussion. K. trinidadensis was described by van den Bold (1958) from Oligocene to Middle
Miocene deep water sediments of Trinidad. It is regarded as the typical form of a large group of
deep-water Krithe species, which are cosmopolitan in distribution but which are especially well
documented from the Caribbean, North Atlantic and southern Europe. The species and subspecies

COLES ET AL.\ OSTRACOD GENUS KRITHE

91

of the K. trinidadensis group vary greatly in size, the observed length range being 0-50-1 -22 mm.
There is also considerable variation in length : height ratio, RPC and anterior vestibulum. However,
the species and subspecies of the K. trinidadensis group are united in sharing the following
morphological characteristics:

1. General shape and outline, particularly the convex dorsum, which is strongly influenced by
pronounced sexual dimorphism.

2. Marked sexual dimorphism; males are relatively and absolutely more elongate than females in
contemporaneous populations, are less inflated posteriorly, have a less strongly convex dorsum,
and a more tapered posterior.

3. Wide inner lamella, with a sinuous line of concrescence and curved VRPC.

4. ADRPC type 3A, with AD 1-3 short, AD 4 elongate, AD 5 very short or absent.

5. Normal overlap of valves (LV > RV).

In any sample or suite of samples of a particular age and locality, the species of this group can
be clearly identified. However, these distinctions frequently break down when specimens from a
wide temporal and spatial range are considered, due to such phenomena as phylogenetic size
variation, size increase with depth and other ecological factors. The species and subspecies which
are considered to belong to the K. trinidadensis group are listed below. Some species are
questionably included, usually due to poor preservation or lack of reliable illustrations of the
internal features given by the original or subsequent authors. Some species among the list will be
shown later to be junior synonyms e.g. K. rex , Dingle el a/., 1990. Valid species or subspecies are
marked with an asterisk *. Frequently a species is based on only male, female or even juvenile
specimens; this is indicated in brackets in each case.

IK. angusta Deltel, 1964 (M), *K. aquilona sp. nov. (M, F), *K. morkhoveni morkhoveni van den
Bold, 1946 (M, F), *K. morkhoveni lamellata ssp. nov., *K. morkhoveni ayressi ssp. nov. (M, F),
K. cancuenensis van den Bold, 1946 (M), K. cancuenensis ambigua Pokorny, 1980 (M), K. citae Oertli,
1961 ( M ), ? AT. contract a Oertli, 1961 (M), *K. crassicaudata van den Bold, 1946 (M, F), K. kollmanni
Pokorny, 1980 (M), K. langhiana Oertli, 1961 (F), K. luyensis Deltel, 1964 (F), IK. peypouqueti
Dingle et al., 1990 (juv), *K praemorkhoveni sp. nov. (M, F), K. elongata van den Bold, 1960 (M)
( = K. prolixa van den Bold, 1966), K. rex Dingle et cd ., 1990 (M, F), *K. trinidadensis van den Bold,
1958 (M, F), K. undecemradiata Ruggieri, 1977 (F), K. sp. D Peypouquet, 1977 (F), K. sp. E
Peypouquet, 1977 (F).

Krithe crassicaudata van den Bold, 1946

Plate 3, figures 6-10; Text-figure 4a-c

1946 Krithe crassicaudata van den Bold, p. 78, pi. 7, fig. 2 a-f

1946 Cytheridea (? Dolocytheridea) guanjayensis van den Bold, p. 83, pi. 7, fig. I ba d.

1960 Krithe crassicaudata van den Bold; van den Bold, p. 158, pi. 3, fig. la-d.

1969 Messinella guanajayensis (van den Bold); van den Bold, p. 396, pi. 1, fig. 1 a-c.

71969 Messinella jamaicensis van den Bold, p. 400, pi. 1, figs 1 a-d. 3a-b, 5 a-b.

1981 Messinella guanajayensis (van den Bold); van den Bold, p. 70, pi. I, fig. 9.

1984 Messinella guanajayensis (van den Bold); Stemeck et ah , p. 1472, fig. 8l.

71985 Krithe sp. 7 Coles, p. 89, pi. 5, figs 23-24; pi. 19, fig. 7.

1987 Messinella jamaicensis van den Bold; Whatley and Coles, p. 88, pi. 2, figs 1-2.

1987 Krithe sp. Cronin and Compton-Gooding, pi. 6, figs 4, 7.

Material and distribution. Forty nine adults and more than 250 juveniles. Hole 549 (M. Eocene), Hole 549A
(U. Eocene-L. Oligocene), Hole 558 (L. Oligocene-L. Miocene).

Dimensions (mm).

FLV GC/NA/951 549A 16-2 L. Oligocene
FLV GC/N A/952 549A 16-2 L. Oligocene
MLV GC/N A/953 549A 13-2 L. Oligocene
FRV GC/NA/954 549A 18-1 U. Eocene
FRY GC/N A/955 549A 39-2 U. Eocene

L

H

0-75

0-50

0-71

0-46

0-80

0-44

0-76

0-45

0-77

0-46

92

PALAEONTOLOGY, VOLUME 37

text-fig. 4. For legend see opposite.

COLES ET AL.: OSTRACOD GENUS KRITHE

93

Dimensions (mm).

MRV GC/NA/956 549A 13-2 L. Oligocene
MRV GC/NA/957 549A 39-2 U. Eocene
JLV GC/N A/958 549A 39-2 U. Eocene
JRV GC/NA/959 549A 39-2 U. Eocene

L

H

0-80

0-44

0-91

0-45

0-66

0-46

0-62

0-40

Stratigraphical range. Middle Eocene to Lower Miocene (NP 18, 20-23; NN 1, ?2).

Diagnosis. A large, thick-shelled Krithe species of the K. trinidadensis group with a pointed postero-
ventral corner. Female subovate, male elongate-subovate, dimorphism not strongly pronounced.
Inner lamella wide ventrally with very sinuous line of concrescence, but narrow postero-dorsally.
Anterior vestibulum small, ‘T’ to ‘Y’-shaped with indented distal margin; ADRPC long with
prominent false ARPC. AD 4 elongate but occasionally short DA 5 short or absent. Posterior
margin bluntly truncate with acute indentation. RV hinge bar posteriorly denticulate.

Remarks. This species is fairly common in the Upper Eocene to Lower Oligocene of Hole 549A but
is rare in Hole 558. The two adult valves from the Lower Miocene of Hole 558 are conspecific with
Krithe sp. 7 of Coles (1985) from the Upper Miocene and Lower Pliocene of the North Atlantic.
These specimens are very similar in shape to K. crassicaudata and may be conspecific although they
have a narrower inner lamella and branching ARPC.

The original illustrations of this species are very poor but the authors have confirmed the identity
of their North Atlantic specimens by examining the type material. This species has been recorded
from the Middle Eocene to Oligocene of Barbados (Steineck et al. 1984), Middle and Upper Eocene
of deep-water deposits in the Caribbean (van den Bold 1960, 1977), Middle Eocene of the NW
Atlantic off New Jersey (Cronin and Compton-Gooding 1987) and the Lower Miocene of Jamaica
(Steineck 1981).

K. crassicaudata probably evolved from K. praemorkhoveni in the late Middle Eocene (NP 18),
although there is a short gap between the ranges of the two species in Hole 549. K. crassicaudata
is larger and more angular in outline than K. praemorkhoveni , and has a sharper postero-ventral
corner. K. crassicaudata is thicker-shelled than K. morkhoveni morkhoveni , and has a sharper
postero-ventral corner and deeper posterior indentation. Both sexes are less elongate than K.
morkhoveni morkhoveni and sexual dimorphism is less marked. The juveniles of K. crassicaudata can
be identified by their thick-shell, subovate shape, and in, the last two instar stages, small anterior
and posterior hinge teeth. These juveniles were described as a new genus, Messinella , by van den
Bold (1969). Although van den Bold recognized that his new genus had ‘some affinity to the
Krithinae' due to the small numbers of false RPC, he considered that the hinge was ‘different from
anything reported in this subfamily’. From the present material, it is clear that Messinella represents

text-fig. 4. a-c. K. crassicaudata van den Bold, a, GC/NA/951 ; DSDP Site 549A, c.c. 16-2; early Oligocene;
female left valve, b, GC/NA/955; DSDP Site 549A, c.c. 39-2; late Eocene; female right valve, c, GC/NA/957;
DSDP Site 549A, c.c. 39-2; late Eocene; male right valve, d-h. K. morkhoveni morkhoveni van den Bold.
D, GC/NA/965 ; DSDP Site 549A, c.c. 8-5 ; late Oligocene ; female left valve, e, GC/NA / 1 08 ; DSDP Site 606A,
c.c. 14; Early Pliocene; female left valve, f, GC/NA/114; DSDP Site 606A, c.c. 15; Early Pliocene; female
right valve. G, GC/NA/102; DSDP Site 606A, c.c. 14; Early Pliocene; male left valve, h, GC/NA/51 ; DSDP
Site 606A, c.c. 14; Early Pliocene; male right valve, i-k. K. morkhoveni lamellalata subsp. nov. i, OS 13976;
DSDP Site 608, c.c. 15; Quaternary; female left valve, j, OS 13980; DSDP Site 607, c.c. 17; Early Pliocene;
female right valve, k, OS 13981; DSDP Site 549A, c.c. 9-5; late Oligocene; female right valve, l-o.
K. morkhoveni ayressi subsp. nov. l, OS 13963; DSDP Site 606, c.c. 9; Late Pliocene; male left valve, m, OS
13974; DSDP Site 606A, c.c. 6; Quaternary; male right valve, n, OS 13972; DSDP Site 558, c.c. 14-1 ; Early
Miocene; male right valve, o, OS 13986; DSDP Site 558, c.c. 13-5; Middle Miocene; female right valve, p-t.
K. trinidadensis van den Bold, p, GC/NA/ 128; DSDP Site 606A, c.c. 13; Late Pliocene; female right valve.
Q, GC/NA/ 127; DSDP Site 606A, c.c. 7; Late Pliocene; female left valve, r, GC/NA/ 126; DSDP Site 606A,
c.c. 14; Early Pliocene; male left valve, s, GC/N A/981 ; DSDP Site 549A, c.c. 9-5; late Oligocene; female right
valve, t, GC/NA/256; DSDP Site 608A, c.c. 13; Early Pliocene ; female left valve. All projectina drawings; x 50.

94

PALAEONTOLOGY, VOLUME 37

the juveniles of K. crassicaudata due to their consistent co-occurrence. The adults resemble the
juveniles in shape and in shell thickness, but have a more angular postero-ventral margin. Only the
posterior of the RV hinge bar is denticulate; the median hinge denticles of the juveniles are lost in
subsequent ontogeny.

We are certain that M. guanajayensis from the Oligocene of Cuba and M. janiaicensis from the
Upper Miocene to Pleistocene of Jamaica are juveniles of K. crassicaudata. Although we have not
seen Plio-Pleistocene material of Messinella , we suspect that this will also prove to be juvenile
specimens of Krithe. In the light of this, we regard Messinella as a junior synonym of Krithe.

The youngest confirmed records of K. crassicaudata are from the Lower Pliocene (NN 15) of
DSDP Sites 609 and 61 1 from the North Atlantic, recorded as the juveniles of Krithe sp. 7 of Coles
(1985). Other records of K. crassicaudata reported as Messinella are as follows: Middle Miocene to
Lower Pliocene of the North Atlantic (Whatley and Coles 1987), Middle Miocene to Pliocene of the
Caribbean (van den Bold 1977), Lower to Middle Miocene of Jamaica (Steineck 1981) (all as
M. janiaicensis), and the Upper Eocene to Lower Miocene of the Caribbean (van den Bold 1977),
Lower Miocene of Jamaica (Steineck 1981 ), Middle Eocene to Oligocene of Barbados (Steineck et al.
1984), Lower Miocene of Haiti (van den Bold 1981) and Lower Oligocene of the Equatorial Pacific
(Steineck et al. 1988) (all as M. guanajayensis).

Krithe morkhoveni morkhoveni van den Bold, 1960
Plate 3, figures 11-18; Text-figure 4d-h

1958 Krithe aff. K. producla Brady; van den Bold, p. 18, pi. 2, figs a, c-d (females).

1960 Krithe morkhoveni van den Bold, p. 160, pi. 3, fig. 6 (female).

1960 Krithe elongata van den Bold, p. 159, pi. 3, fig. 5 a-e (males) [junior homonym of K. elongata
Jones and Kirkby, 1898].

1961 Krithe langhiana Oertli, p. 24, pi. 3, figs 24-30 (females).

1961 Krithe citae Oertli, p. 25, pi. 3, figs 21-32 only (males).

1961 Krithe contracta Oertli; p. 26, pi. 3, figs 35-38 (juveniles).

1964 Krithe luyensis Deltel, p. 171, pi. 4, figs 83-85 (females).

1964 Krithe compressa dertonensis Ruggieri; Dieci and Russo, p. 79, pi. 15, fig. 7 (male).

1964 Krithe cf. contracta Oertli; Dieci and Russo, p. 79, pi. 15, fig. 8 (female).

1964 Xestoleberis subtruncata Dieci and Russo, p. 85, pi. 2, fig. 13 a-b; pi. 16, fig. 7 (females).

1966 Krithe prolixa van den Bold, p. 180 (males) [new name for K. elongata , van den Bold non Jones].

1967 Krithe aff. morkhoveni van den Bold; Ascoli, p. 54, pi. 1, figs 4-6 (females).

1968 Krithe aff. morkhoveni van den Bold; Russo, p. 39, pi. 6, figs 4 a, d (female); pi. 8, fig. 6; pi. 9,

figs 1«, c (males).

1968 Krithe trinidadensis van den Bold; van den Bold, pi. 2, figs 10 a-b (male), 10 c-d (female).

1972 Krithe langhiana Oertli; Sissingh, p. 171, pi. 4, figs 6 a-b (female).

1974 Krithe aff. K. bartonensis (Jones); Leroy and Levinson, p. 24, pi. 11, fig. 4?; pi. 12, fig. 5 (males).

1974 Krithe undecemradiata Ruggieri, p. 176, figs 6, 3 a-b (male); 4 a-b (female).

1977 Krithe sp. D llfn Peypouquet, p. 113, fig. 37 [pars ] (female).

1977 Krithe sp. D 12fn Peypouquet, p. 113, fig. 37 [pars ] (female).

1977 Krithe sp. D22 Peypouquet, p. 113, fig. 37 [pars] (female).

1980 Krithe cancuensis ambigua Pokorny, p. 342, figs 8-10; pi. 2, figs 2-3 (males).

1980 Krithe kollmanni Pokorny, p. 338, figs 1-3; pi 1, figs 1-3; pi. 2, fig. 1 (females).

1981 Krithe prolixa van den Bold; Steineck, p. 359; pi. 1, fig. 13 (males).

1981 Krithe sp. 6 Ciampo, p. 67, pi. 6, fig. 3 (female).

1981 Krithe sp. 8 Ciampo, p. 67, pi. 6, fig. 4 (female).

1981 Krithe sp. 5 Ciampo, p. 67, pi. 6, fig. 7 (male).

1984 Krithe morkhoveni van den Bold; Steineck et al., p. 1473, figs e, i (females).

1984 Krithe prolixa van den Bold; Steineck et al., fig. j (males).

1985 Krithe luyensis Deltel; Ducasse et al., p. 285, pi. 78, figs 1 1-13 (females).

Material and distribution. Three hundred and fifty adult valves. Hole 549 (U. Palaeocene-U. Eocene), Hole
549A (U. Eocene-U. Oligocene, U. Miocene), Hole 558 (L. Oligocene-U. Miocene), Hole 563 (M.-U.

COLES ET AL.\ OSTRACOD GENUS KRITHE

95

Miocene), Holes 606, 606A, 610 (Pliocene-Quaternary), Holes 607, 608 (U. Miocene-Quaternary), Holes 609B
(U. Pliocene-Quaternary), Holes 609, 609C (U. Pliocene), Hole 610E (U. Miocene-L. Pliocene), Hole 61 ID
(Pliocene).

Dimensions (mm). L H

FLV GC/NA/963 549 H3-2 U. Eocene

0-83

0-47

FLV GC/N A/964 549A 8-5 U. Oligocene

0-78

0-43

FLV GC/NA/965 549A 8-5 U. Oligocene

0-70

0-40

FLV GC/NA/108 606A-14 L. Pliocene

0-98

0-58

FLV GC/NA/121 606A-17 L. Pliocene

0-91

0-54

MLV GC/N A/966 549 15-5 L. Eocene

0-82

0-38

MLV GC/NA/967 549A 8-5 U. Oligocene

0-87

0-43

MLV GC/NA/1 18 606A-14 L. Pliocene

104

0-40

MLV GC/NA/1 38 606-5 Quaternary

0-98

0-47

MLV GC/NA/102 606A-14 L. Pliocene

0-94

0-44

FRV GC/NA/968 549A 8-5 U. Oligocene

0-76

0-49

FRV GC/NA/969 549 H3-2 U. Eocene

0-84

0-45

FRV GC/NA/970 549A 8-5 U. Oligocene

0-67

0-37

FRV GC/NA/1 14 606A-15 L. Pliocene

0-91

0 51

MRV GC/NA 971 549A8-5 U. Oligocene

0-88

0-39

MRV GC/N A/51 606A-14 L. Pliocene

103

0-40

MRV GC/NA/44 606A-14 L. Pliocene

0-91

0-40

Stratigraphical range. Upper Palaeocene to Recent.

Diagnosis. A medium (females only) to large (females and males), subovate (females) to elongate
subrectangular (males) subspecies of Krithe. Dorsum convex, most strongly so in females. Greatest
height at mid-length (females) or anterior of mid-length (males). Inner lamella moderately wide with
sinuous line of concrescence. Anterior vestibulum small, ‘mushroom '-shaped. ADRPC type 3A.

Remarks. This is an abundant subspecies in the Lower Eocene to Quaternary of the North Atlantic.
One specimen from the Upper Palaeocene of Hole 549 is thicker-shelled than the Eocene to Miocene
specimens, and may be a post-maturation moult of K. praemorkhoveni. There is a considerable
degree of variation within this subspecies, as shown by the specimens figured in this study. This
variation is expressed in the following features.

1. Size. In this study the length range of this species is 0-67 to 0-87 mm (23 per cent range).

2. Shape. Some variation in the L:H ratio, valve inflation and outline occurs, particularly in the
angle of the postero-dorsal slope.

3. Inner lamella. This varies in width, particularly in the posterior region. A few specimens
approach the form of K. morkhoveni lamellalata subsp. nov.

4. ADRPC pattern. The length and position of the ADRPC is variable; in particular AD 1-3 may
be very short where the inner lamella is narrow, AD 4 may arise from the vestibular neck or from
the antero-dorsal fused zone and AD 5 may be very short or apparently absent.

K. morkhoveni morkhoveni is thought to have evolved from K. praemorkhoveni , probably in the
early Eocene, by becoming larger, thinner-shelled, and developing a wide inner lamella.
K. trinidadensis evolved from K. morkhoveni morkhoveni in the Middle Eocene of the Atlantic in the
area of Hole 549, while the subspecies K. morkhoveni ayressi evolved from the nominate subspecies
in the Lower Oligocene (see below).

K. morkhoveni morkhoveni is a widespread deep-water subspecies, recorded by numerous authors
under a variety of names and there have been many errors of identification such as its record as
K. compressa dertonensis Ruggieri by Dieci and Russo (1964) and as K. bartonensis (Jones) by Leroy
and Levinson (1974); in both cases the material differed widely from the species to which they were
assigned.

96

PALAEONTOLOGY, VOLUME 37

In addition to the North Atlantic occurrences cited above, the following are noted: Eocene and
Oligocene of Aquitaine (Deltel 1964; Ducasse et al. 1985), Eocene and Ohgocene of the Bay of
Biscay and Eocene of the Rockall Plateau (among the material of Ducasse and Peypouquet 1979
seen by GC), Oligocene of Sicily (Ciampo 1981), Upper Eocene and Lower Oligocene of the Angola
Abyssal Plain in the South Atlantic and Upper Eocene of Denmark (GC personal observations).
Middle and Upper Miocene of northern Italy (Oertli 1961; Died and Russo 1964, 1967; Ascoli
1968), Upper Miocene of Crete (Sissingh 1972), Upper Eocene to Pliocene of the Caribbean region
(van den Bold 1977, 1981), Pleistocene of Bologna, Italy (Ruggieri 1974), Pleistocene of the Gulf
of Mexico (Leroy and Levinson 1974), Pleistocene of the Iberian Portal at 2798 m (Elant 1985) and
at 2421 and 3700 m (Harpur 1985), Pleistocene of the North Atlantic in cores L4, L5, N3, P2, R2,
R3, R4, R5, S3 and T3 between 1678 and 4566 m PDWD and lat. 43° to 68° N (among the material
of Porter 1984 seen by GC) and Quaternary of the southwestern Pacific (Ayress 1988).

In the Recent this species occurs in the Atlantic off Florida between 739 and 1029 m (among the
material of Cronin seen by GC), west of the Iberian Portal at 2798 m (Elant 1985) and in the North
Atlantic between 1775 to 5440 m (material of Barkham 1985 and personal observations). It is most
abundant below 2000 m and appears to be absent from the present day Mediterranean; indicating
that this subspecies is a reliable indicator of deep oceanic waters for post-Eocene time.

Krithe morkhoveni lamellalata subsp. nov.

Plate 4, figures 1-3; Text-figure 4i-k

Derivation of name. Latin, with reference to the wide calcified inner lamella of this subspecies.

Holotype. Female left valve, OS 13976.

Type locality and horizon. DSDP Site 608, Hole 608, mid-North Atlantic, King’s Trough, northeast of the
Azores. Lat. 42° 0 21 N ; long. 23° 05-25' W. PDWD 3526 m. Core catcher 5, Quaternary, NN 19. Nannofossil
ooze.

Material and distribution. One hundred and nineteen adult valves. Hole 549A (U. Eocene-U. Oligocene,
U. Miocene), Hole 558 (U. Oligocene, M. Miocene), Hole 607 (L. Pliocene), Hole 608 (U. Miocene-Quaternary),
Hole 609 (U. Pliocene), Holes 609B, 610C (Quaternary), Hole 610 (U. Miocene-L. Pliocene), Hole 610D
(U. Miocene), Holes 610E, 61 ID (U. Pliocene).

EXPLANATION OF PLATE 4

Figs 1-3. Krithe morkhoveni lamellalata subsp. nov. 1, OS 13978; DSDP Site 610E, c.c. 6; Late Miocene;
female right valve, external view; x 50. 2-3, OS 13976; DSDP Site 608, c.c. 5; Quaternary; female left valve,
external and internal views; x44.

Figs 4-6. Krithe morkhoveni ayressi subsp. nov. 4, OS 13963; DSDP Site 606, c.c. 9; Late Pliocene; male left
valve, external view; x 53. 5, OS 13969; DSDP Site 558, c.c. 24-3; late Oligocene; female right valve,
external view; x 60. 6, OS 13973; DSDP Site 608, c.c. 5; Quaternary; male left valve, internal view; x 50.

Figs 7- 12. Krithe trinidadensis van den Bold. 7, GC/NA/127; DSDP Site 606A, c.c. 7; Late Pliocene; female
left valve, external view; x31. 8-9, GC/NA/128; DSDP Site 606A, c.c. 13; Late Pliocene; female right
valve, external and internal views; x31. 10-1 1, GC/NA/ 124; DSDP Site 606A, c.c. 14; Early Pliocene; male
right valve, external and internal views; x33. 12, GC/NA/126; DSDP Site 606A, c.c. 14; Early Pliocene;
male left valve, external view; x31.

Figs 13-17. Krithe aquilonia sp. nov. 13-15, OS 13996; DSDP Site 610, c.c. 5; Quaternary; female left valve,
external and internal views; x 66. 16, OS 13999; DSDP Site 610E, c.c. 3; Late Miocene; male left valve,
external view; x 62. 17, OS 14000; DSDP Site 608, c.c. 1 1 ; Late Pliocene; male right valve, external view;
x 62.

All figures are scanning electron micrographs.

PLATE 4

COLES et al., Krithe

98

PALAEONTOLOGY, VOLUME 37

Dimensions (mm).

L

H

Holotype FLV OS 13976 608-5 Quaternary

0-97

0-56

Paratype FLV OS 13977 608-7 U. Pliocene

0-90

0-53

Paratype FLV OS 13979 610E-5 U. Miocene

0-96

0-50

Paratype FRV OS 13981 549A 9-5 U. Oligocene

0-88

0-46

Paratype FRV OS 13980 607-17 L. Pliocene

0-87

0-50

Paratype FRV OS 13978 61CL4 Quaternary

0-93

0-50

Paratype MLV OS 13982 549A 5-5 U. Miocene

0-94

0-47

Paratype MRV OS 13983 549A 5-5 U. Miocene

0-92

0-43

Stratigraphical range. Upper Eocene to Upper Oligocene, Middle Miocene to Quaternary (NP 20-5 ; NN 5, 9,
11-13, 15-19). The subspecies probably also occurs in the Recent of the North Atlantic.

Diagnosis. A large, strongly inflated subspecies of Krithe expressing moderate sexual dimorphism.
Outline very similar to that of K. morkhoveni morkhoveni but rather more elongate-subovate. Inner
lamella very wide, much wider than in K. morkhoveni morkhoveni , especially so posteriorly, with a
highly sinuous line of concrescence, deeply depressed below adductors. Anterior vestibulum small,
elongate 'mushroom' to 'Y '-shaped, with very narrow opening, long narrow neck and branching
distally. Other features as for K. morkhoveni morkhoveni .

Remarks. K. morkhoveni lamellalata differs from the nominative subspecies in its much wider inner
lamella, more sinuous line of concrescence and more elongate-subovate outline. In the present
material, K. morkhoveni lamellalata is common only in the Oligocene and Upper Miocene of Hole
549A. It also occurs in the Pleistocene of the North Atlantic in core L4 at 3422 m and lat. 43° N
(among the material of Porter 1984 seen by GC) and west of the Iberian Portal at 2421 m (along
the material of Harpur 1985 seen by GC).

Krithe morkhoveni ayressi subsp. nov.

Plate 4, figures 4-6; Text-figure 4l-o

Derivation of name. Named for Dr Michael Ayress, in recognition of his work on Krithe and other deep-sea
Ostracoda from the Indo-Pacific.

Holotype. Male left valve OS 13963.

Type locality and horizon. DSDP Site 606, Hole 606, mid-North Atlantic, S.W. of the Azores. Lat. 37° 20-32' N,
long. 35° 29-99' W. PDWD 3007 m. Core catcher 9, Upper Pliocene, NN 18. Nannofossil ooze.

Material and distribution. One hundred and thirty six adult valves. Hole 558 (L. Oligocene-U. Miocene),
Holes 606, 606A (U. Pliocene-Quaternary), Hole 608 (U. Miocene, U. Pliocene-Quaternary, Hole 607
(Pliocene-Quaternary), Holes 610E, 61 ID (L. Pliocene), Holes 609A, 609B, 610 (Quaternary).

Dimensions (mm).

L

H

Holotype MLV OS 13963 606-9 U. Pliocene

0-79

0-36

Paratype FLV OS 13964 558 10-1 M. Miocene

0-72

0-39

Paratype FLV OS 13965 558 14-5 L. Miocene

0-74

0-36

Paratype FLV OS 13966 588 22-1 U. Oligocene

0-75

0-39

Paratype MLV OS 13973 608-5 Quaternary

0-84

0-40

Paratype FRV OS 13967 558 16-1 L. Miocene

0-71

0-36

Paratype FRV OS 13968 558 13-5 M. Miocene

0-71

0-33

Paratype FRV OS 13969 558 24-3 U. Oligocene

0-70

0-34

Paratype MRV OS 13970 558 24-3 U. Oligocene

0-81

0-33

Paratype MRV OS 13971 558 25-5 U. Oligocene

0-84

0-31

Paratype MRV OS 13972 558 14-1 L. Miocene

0-76

0 31

COLES ET A L. : OSTRACOD GENUS KRITHE

99

Dimensions (mm). L H

Paratype MRV OS 13974 606A-6 Quaternary 075 0 35

Paratype MRV OS 13976 607-18 L. Pliocene 0-79 0-35

Stratigraphical range. Lower Oligocene to Quaternary (NP 21?, 22-25; NN 2, 5, 8?, 11-12, 15-21) in the
present study. The species is extant.

Diagnosis. A medium to large, subrectangular, elongate subspecies of Krithe with slight sexual
dimorphism. Straight (male) to slightly convex (female) dorsal margin; posterior bluntly truncate
with very steep postero-ventral slope. Inner lamella moderately wide; anterior vestibulum large,
‘mushroom ’-shaped. RV with secondary accommodation groove.

Remarks. This taxon is considered to be a subspecies of K. morkhoveni because of its close
morphological similarity to the nominative subspecies. However, K. morkhoveni ayressi is smaller
than K. morkhoveni morkhoveni and is more elongate with a less tapered posterior in the males.
Sexual dimorphism is less marked than in K. morkhoveni morkhoveni ; the females of K. morkhoveni
ayressi are elongate-subovate, and are relatively lower and less inflated. In other features such as
ADRPC pattern and inner lamella, the two subspecies are indistinguishable.

In the present material, K. morkhoveni ayressi is only common in the Oligocene of Hole 558, but
occurs persistently if rarely in the Miocene to Quaternary of the North Atlantic. It probably evolved
from K. morkhoveni morkhoveni in the Lower Oligocene. This subspecies has also been found in the
Recent of the Gulf of Mexico at a depth of 1500 m (personal observations), although these
specimens have a smaller anterior vestibulum.

Krithe trinidadensis van den Bold, 1958

Plate 4, figures 7-12; Text-figure 4p-t, Text-figure 5a-b

1958 Krithe trinidadensis van den Bold, p. 398, pi. 1, figs 3 a, 6e (males); 3 c-d, f (females).

1969 Krithe producta Brady; Yassini, pi. 22, fig. 4 (female).

1977 Krithe sp. D21 en Peypouquet, p. 113, fig. 37 [pars] (female).

71977 Krithe sp. D22 en Peypouquet, p. 113, fig. 37 [pars].

1981 Krithe sp. 1 Ciampo, p. 67, pi. 6, fig. 5 (male).

?1990 Krithe peypouqueti Dingle et al ., p. 279, figs 17l-m, 18l, 21e-f, 22a (juveniles).

1990 Krithe rex Dingle et al., p. 276, figs 17g-h, 18h, 21a-d (males and females).

Material and distribution. Three hundred and thirty nine adult valves. Hole 549 (M.-U. Eocene), Hole 549A
(U. Eocene-U. Oligocene, U. Miocene), Hole 558 (L. Oligocene-U. Miocene), Hole 563 (M. Miocene-
U. Miocene), Holes 606, 609B (U. Pliocene-Quaternary), Holes 606A, 608, 608A (Pliocene-Quaternary), Hole
607 (U. Miocene-Quaternary), Holes 609, 609A, 610C (Quaternary), Holes 610, 610D (U. Miocene), Hole
610B (U. Pliocene), Hole 610E (U. Miocene-L. Pliocene), Hole 61 ID (L. Pliocene, Quaternary).

Dimensions (mm).

L

H

FLV GC/NA/979 549 2-4 M. Eocene

0-98

0-60

FLV GC/NA/983 549 H3-2 U. Eocene

0-87

0-58

FLV GC/NA/127 606A-7 U. Pliocene

L36

0-89

FLV GC/NA/161 607-6 Quaternary

L34

0-86

FLV GC/NA/256 608A-13 U. Pliocene

115

0-73

FLV GC/NA/148 606-1 Quaternary

L32

0-82

MLV GC/NA/980 549 5-1 M. Eocene

110

0-56

MLV GC/NA/126 606A-14 L. Pliocene

L36

0-55

MLV GC/NA/159 607-6 Quaternary

1 -57

0-78

FRV GC/N A/984 549A 16-1 L. Oligocene

109

0 61

FRV GC/NA/981 549A 9-5 U. Oligocene

1-22

0-70

FRV GC/NA/146 606-4 Quaternary

L37

0-80

FRV GC/NA/128 606A-13 U. Pliocene

1 34

0-80

100

PALAEONTOLOGY, VOLUME 37

COLES ET A L. : OSTRACOD GENUS KRITHE

101

Dimensions (mm).

FRV GC/NA/162 607-6 Quaternary
MRV GC/NA/982 549 H3-2 U. Eocene
MRV GC/NA/124 606A-14 L. Pliocene
MRV GC/NA/160 607-6 Quaternary
MRV GC/NA/130 606A-5 Quaternary

L

H

L27

0-74

0-99

045

1-26

0-57

1-52

0-70

L58

0-76

Stratigraphical range. Middle Eocene to Quaternary (NP 16-25; NN 1, 3-21) in the present study, also extends
to the present day.

Diagnosis. A very large, thick-shelled species of Krithe with strong sexual dimorphism. Female
subovate and very tumid; male elongate to very elongate, subtriangular and less strongly inflated
than female. Inner lamella broad; anterior vestibulum ‘T’ or "Y ’-shaped to small ‘pocket ’-shaped.
ADRPC type 3A. Hinge strong, LV with arched, deep accommodation groove. Central muscle scars
large, increasing in size dorsally; most dorsal adductor reniform to ‘ U ’-shaped, trefoil frontal scar.

Remarks. K. trinidadensis was described by van den Bold (1958) from the Oligocene and Miocene
of Trinidad. It is very similar in shape, inner lamella and ADRPC pattern to its probable ancestor,
K. morkhoveni morkhoveni , but is always much larger and more inflated; the A-l juveniles of
K. trinidadensis are approximately equal in size to the adult females of K. morkhoveni morkhoveni .

The anterior vestibulum ranges from a large ‘ T ’ to ‘ Y ’ shape to a small ‘ pocket ’ shape and, while
most specimens have either small or large vestibula, the present material includes specimens with
an anterior vestibulum intermediate in form between the two types (e.g. GC/NA/980). There are
few published records or illustrations of this species, probably because it has been included within
K. morkhoveni and other species. In addition to the occurrences in the present study material, the
following are noted: Upper Oligocene of Sicily (Ciampo 1981), Upper Miocene to Quaternary of
the Rockall Plateau (among the material of Ducasse and Peypouquet 1979 seen by GC), Pleistocene
of the Iberian Portal at 2798 m (Elant 1985) and at 2414 and 3700 m (Harpur 1985) and Pleistocene
of the North Atlantic in cores L4, P2, R2, R3, R4 and R5 between 1678 and 3422 m and lat. 43°
to 61° N (material of Porter 1984 seen by GC). In the Recent it ranges in the North Atlantic from
1320 to 5440 m (Davies 1981 ; Barkham 1985; GC, RCW, personal observations), and also occurs
at 1080 m off Florida (material of Cronin 1983 seen by GC) at 1500 m in the Gulf of Mexico

text-fig. 5. a-b. K. trinidadensis van den Bold, a, GC/NA/130; DSDP Site 606A, c.c. 5; Quaternary; male
right valve, b, GC/NA/160; DSDP Site 607, c.c. 6; Quaternary; male right valve, c-f. K. aquilonia sp. nov.
c, OS 13996; DSDP Site 610, c.c. 5; Quaternary; female left valve, d, OS 13998; DSDP Site 608, c.c. 5;
Quaternary; female right valve, e, OS 13999; DSDP Site 620E, c.c. 3; Late Miocene; male left valve.
f, OS 14000; DSDP Site 608, c.c. 11; Late Pliocene; male right valve, g— j. Krithe praemorkhoveni sp. nov.
G, OS 13984; DSDP Site 549, c.c. 16-3; late Palaeocene; female right valve, h, OS 13985; DSDP Site 549, c.c.
14-4; early Eocene; female left valve. I, OS 13988; DSDP Site 549, c.c. 8-1 ; middle Eocene; male left valve.
J, OS 13993; DSDP Site 549, c.c. 16-3; late Palaeocene; male right valve, k-q. K. pernoides pernoides
(Bornemann). k, GC/NA/310; DSDP Site 61 ID, c.c. 12; Late Pliocene; female left valve. L, GC/NA/844;
DSDP Site 549A, c.c. 8-5; late Oligocene; female left valve, m, GC/NA/171 ; DSDP Site 607, c.c. 10; Late
Pliocene; female right valve, n, GC/NA/311 ; DSDP Site 61 ID, c.c. 12; Late Pliocene; female right valve,
o, GC/NA/849; DSDP Site 549A, c.c. 8-5; late Oligocene; male left valve, p, GC/NA/846; DSDP Site 549,
c.c. 2-1 ; middle Eocene; female right valve. Q, GC/NA/850; DSDP Site 5409 A, c.c. 8-5; late Oligocene; male
right valve. R-v. K. pernoides sinuosa Ciampo. r, GC/NA/852; DSDP Site 563, c.c. 10-1 ; Middle Miocene;
female right valve, s, GC/NA/35; DSDP Site 606A, c.c. 5; Quaternary; female right valve. T, GC/NA/180;
DSDP Site 607, c.c. 13; Late Pliocene; female left valve, u, GC/NA/156; DSDP Site 607, c.c. 5; Quaternary;
male right valve, v, GC/NA/155; DSDP Site 607, c.c. 5; Quaternary; male left valve, w-z. Krithe sp. cf.
parvula Deltel. w, GC/NA/926; DSDP Site 549, c.c. 11-1; early Eocene; female left valve, x, GC/NA/928;
DSDP Site 549, c.c. 8-4; middle Eocene; female right valve. Y, GC/NA/932; DSDP Site 549, c.c. 16-3; Late
Pliocene; male left valve, z, GC/NA/931; DSDP Site 549, c.c. 11-1; early Eocene; male right valve.
aa-bb. Krithe sp. 1. aa, GC/NA/892; DSDP Site 549, c.c. 19-3; Late Pliocene; female? left valve, bb,
GC/NA/893; DSDP Site 549, c.c. 19-3; Late Pliocene; female? right valve. All projectina drawings; x 50.

102

PALAEONTOLOGY, VOLUME 37

(GC personal observations) and in the Bay of Biscay at bathyal and abyssal depths down to 3950 m
(Yassini 1969, as K. product a Brady). It has been described as a new species, K. rex , by Dingle et at.
(1990) from the South Atlantic off southwestern Africa at a depth of 2916 m, while another of
their species, K. peypouqueti, recorded at 2916 m and 4736 m in the same area probably represents
the juveniles of this species.

Krithe aqui Ionia sp. nov.

Plate 4, figures 13-17; Text-figure 5c-f

Derivation of name. Latin, with reference to the northerly distribution of this species.

Holotvpe. Female left valve OS 13996.

Type locality and horizon. DSDP Site 610, Hole 610, eastern North Atlantic, west of the Rockall Trough on
the Feni Drift. Lat. 53° 13 30' N, long. 18° 53 21 W. PDWD 2417 m. Core catcher 5, Quaternary, NN 19.
Nannofossil ooze.

Material and distribution. Fifty three adult valves. Hole 558 (L. Miocene), Holes 607, 609, 61 ID (L. Pliocene),
Hole 608 (Pliocene-Quaternary), Holes 608A, 610B (U. Pliocene), Hole 609B (U. Pliocene-Quaternary), Hole
610 (U. Miocene, Quaternary), Hole 610C (Quaternary), Hole 610D (U. Miocene), Hole 610E (U. Miocene-
L. Pliocene).

Dimensions (mm).

L

H

Holotype FLV OS 13996 610-5 Quaternary

0-65

0-36

Paratype FLV OS 13997 08-11 U. Pliocene

0-63

0-33

Paratype FRV OS 13998 608-5 Quaternary

0-63

0-31

Paratype MLV OS 13999 610E-3 U. Miocene

0-67

0-31

Paratype MRV OS 14000 608-11 U. Pliocene

0-67

0-30

Paratype MRV OS 14001 558 14-1 L. Miocene

0-62

0-28

Stratigraphical range. Miocene-Quaternary (NN 1, 4, 5, 10-21) in

the present study.

It almost certainly also

occurs in the Recent.

Diagnosis. A small, subrectangular, elongate species of Krithe with moderate sexual dimorphism.
Females relatively higher and more inflated than elongate males. Both sexes with gently (LV) to
quite strongly convex (RV) dorsum, and bluntly truncate posterior. Inner lamella wide, anterior
vestibule small ‘T’ to ‘mushroom ’-shape. ADRPC type 3A.

Description. Small, elongate, subrectangular and strongly inflated, especially so in female. FLV dorsal margin
regularly convex, continuously curved with broadly rounded anterior; ventral margin with very shallow oval
incurvature at mid-length; posterior bluntly truncate, very steep postero- ventral slope at about 85° to the
ventral margin. FRV as FLV but with a very slight antero-dorsal concavity. Male similar, but more elongate
and dorsal margin straighter, especially so in MLV. In both sexes, selvage narrow, slight posterior indentation
and shallow posterior pit; moderately thin-shelled. LV overlaps RV and overreaches RV along dorsal margin.
NPC small, regularly distributed. Inner lamella wide, deeply depressed below adductors. Anterior vestibule
small, ‘T‘ to ‘mushroom ’-shaped, with narrow opening and long, narrow neck. Ten to eleven short, straight
ARPC in fan arrangement; ADRPC type 3A; three to four short to moderately elongate PDRPC. Posterior
vestibule similar in size to anterior vestibule, elongate subrectangular, with up to five very short PRPC. Hinge
pseudodont; RV with slightly arched hinge bar which is raised posteriorly and bears up to six tiny denticles,
complementing posteriorly locellate groove in LV. Central muscle scars relatively large, consisting of a slightly
arcuate row of four adductors, the topmost is reniform and the middle scars are biconcave. Trefoil frontal scar
and ovate mandibular scar.

Remarks. K. aquilonia is distinctly smaller than K. morkhoveni ayressi , which may be its ancestor.
This species is also recorded from the Upper Miocene of the Rockall Trough (among the material
of Ducasse and Peypouquet 1979 seen by GC) and in the Pleistocene of the North Atlantic between

COLES ET AL.\ OSTRACOD GENUS KRITHE

103

lat. 43° and 61° N and PDWD 938 to 4566 m in cores L4, L5, Ol, P2, R2 and R3 of Porter ( 1984).
K. aquilonia is only known from the Miocene to Quaternary of the North Atlantic between lat. 37°
and 61° N, with no records from other oceans.

Krithe praemorkhoveni sp. nov.

Plate 5, figures 1-6; Text-figure 5g-j

Derivation of name. Latin, referring to the ancestral relationship of this species to K. morkhoveni.
Holotvpe. Female right valve OS 13984.

Type locality and horizon. DSDP Site 549, Hole 549, Goban Spur. Lat. 49° 05 28' N, long. 13° 05-88' W;
PDWD 2513 m. Core 16, section 3, interval 0-41-0-48 m. Upper Palaeocene, NP 9. Olive-grey nannofossil
chalk.

Material and distribution. One hundred and ninety five adult valves. Hole 549 (U. Palaeocene-M. Eocene),
Hole 550 (L. Palaeocene-L. Eocene).

Dimensions (mm).

Holotype FRV OD 13984 549 16-3 U. Palaeocene
Paratype FLV OS 13985 549 14-4 L. Eocene
Paratype FLV OS 13986 54915-5 L. Eocene
Paratype MLV OS 13987 549 11-1 L. Eocene
Paratype MLV OS 13988 549 8-1 M. Eocene
Paratype MLV OS 13989 549 13-4 L. Eocene
Paratype FRV OS 13990 549 11-1 L. Eocene
Paratype FRV OS 13991 549 9-2 M. Eocene
Paratype FRV OS 13992 549 15-5 L. Eocene
Paratype MRV OS 13993 549 16-3 U. Palaeocene
Paratype MRV OS 13994 549 11-1 L. Eocene
Paratype MRV OS 13995 549 15-5 L. Eocene

L

H

0-58

0-34

0-54

0-35

0-59

0-37

0-64

0-34

0-68

0-33

0-60

0-33

0-57

0-34

0-71

0-41

0-59

0-36

0-66

0-32

0-62

0-30

0-66

0-34

Stratigraphical range. Lower Palaeocene to Middle Eocene (NP 3—4, 6-16) in the present study.

Diagnosis. A medium, thick-shelled, strongly sexually dimorphic species of Krithe belonging to the
K. trinidadensis group, similar to K. morkhoveni van den Bold, but smaller, more robust and with
a narrower inner lamella.

Description. Medium-sized, thick-shelled and moderately inflated. Females subovate, males subovate to
elongate-subrectangular. FLV dorsum strongly convex, anterior broadly rounded, posterior bluntly truncate,
ventral margin convex. FRV as FLV but dorsum more convex, with shallow antero-dorsal concavity and
slightly convex ventral margin. Males similar but more elongate, with less convex dorsum; ventral margin may
be concave. Normal LV > RV overlap, NPC relatively large. Inner lamella moderate width with sinuous line
of concrescence. Anterior vestibulum small, ‘mushroom’ to ‘T ’-shaped, with eleven to twelve ARPC. ADRPC
type 3A; AD 1-3 very short, AD 4 elongate and arises from the vestibular neck. Hinge adont; frontal scar may
be quadrifoil.

Remarks. This is the most abundant Krithe species in the Palaeocene and Lower Eocene of Holes
549 and 550, but does not occur above the Middle Eocene (NP 16). It is the ancestor of
K. morkhoveni morkhoveni , from which it differs in being smaller (length range 0-54 to 0-71 mm) and
thicker-shelled.

The Lower Eocene species K. kollmanni Pokorny, 1980, and K. cancuenensis ambigua Pokorny,
1980, from deep-water Globigerina marls in the former Czechoslovakia, probably represent the
female and male, respectively of the present species. However, the early Oligocene holotypes of both
species are larger and thinner-shelled and are included within K. morkhoveni morkhoveni.

104

PALAEONTOLOGY, VOLUME 37

ADRPC TYPE 3A

(Other species not assigned to K. trinidadensis Group)

Krithe pernoides pernoides (Bornemann, 1855)

Plate 5, figures 7-12; Text-figure 5k-q

71855 Bairdia pernoides Bornemann, p. 358, pi. 20, fig. la-c
1918 Krithe pernoides (Bornemann); Kuiper, p. 36, pi. 1, figs 12a, c? only.

1957 Krithe pernoides (Bornemann); Keij, p. 86, pi. 6, fig. I i a-b.

1962 Krithe pernoides (Bornemann); Bassiouni, p. 22, pi. 9, figs 1-3.

1969 Krithe pernoides (Bornemann); Schremeta, p. 90, pi. 7, figs 8-10.

1969 Krithe pernoides (Bornemann); Pietrzeniuk, p. 24, pi. 5, fig. 117; pi. 15, figs 13, 14?

Material and distribution. Two hundred and ninety five adult valves. Hole 549 (M.-U. Eocene), Hole 549A
(U. Eocene-U. Oligocene), Hole 558 (L. Oligocene-M. Miocene), Hole 607 (U. Pliocene-Quaternary), Hole
608A (Quaternary).

Dimensions ( mm ).

L

H

FLV GC/NA/843 549A 16-2 L. Oligocene

0-59

0-33

FLV GC/NA/844 549A 8-5 U. Oligocene

0-59

0-34

FLV GC/NA/884 549A 8-5 U. Oligocene

0-59

0-34

FLV GC/NA/255 608A-3 Quaternary

0-73

0-43

FRV GC/NA/845 549A 8-5 U. Oligocene

0-58

0-30

FRV GC/NA/846 549 2-1 M. Eocene

0-57

0-29

FRV GC/NA/171 607-10 U. Pliocene

0-73

0-40

MLV GC/N A/847 549 5-1 M. Eocene

0-64

0-29

MLV GC/NA/848 549A 16-1 L. Oligocene

0-81

0-39

MLV GC/NA/849 549A 8-5 U. Oligocene

0-70

0-34

MRV GC/NA/850 549A 8-5 U. Oligocene

0-67

0-29

FLV GC/NA/3 10 6 1 1 D 1 2 L. Pliocene

0-78

0-43

FRV GC/NA/31 1 61 ID 12 L. Pliocene

0-75

0-38

Stratigraphical range. Middle Eocene to Middle Miocene, Upper Pliocene to Quaternary (NP 15-25; NN 4-5,
18-19).

Diagnosis. A medium (female) or medium to large (male) subspecies of Krithe with an elongate,
subrectangular carapace, subparallel dorsal and ventral margins, wide inner lamella, long RPC, and
ADRPC type 3A. Posterior convex, with no posterior concavity and no marked posterior angle.

EXPLANATION OF PLATE 5

Figs 1-6. Krithe praemorkhoveni sp. nov. 1-2, OS 13986; DSDP Site S49, c.c. 15-5; early Eocene; female left
valve, external and internal views; x 71. 3, OS 13992; DSDP Site S49, c.c. 15-5; early Eocene; female right
valve, external view; x71. 4, OS 13989; DSDP Site S49, c.c. 13-4; early Eocene; male left valve, external
view; x 70. 5-6, OS 13995; DSDP Site S49, c.c. 15-5; early Eocene; male right valve, external and internal
views; x64.

Figs 7-12. Krithe pernoides pernoides (Bornemann). 7, GC/NA/884; DSDP Site S49A, c.c. 8-5; late Oligocene;
female left valve, external view; x71. 8, GC/NA/255; DSDP Site 608A, c.c. 3; Quaternary; female left
valve, internal view; x 57. 9-10, GC/NA/171 ; DSDP Site 607, c.c. 10; Late Pliocene; female right valve,
external and internal views; x 57. 1 1-12, GC/NA/850; DSDP Site 549A, c.c. 8-5; late Oligocene; male right
valve, external and internal views; x63.

Figs 13-17. Krithe pernoides sinuosa Ciampo. 1 3-14, GC/NA/35 ; DSDP Site 606A, c.c. 5; Quaternary ; female
right valve, external and internal views; x 56. 1 5-16, GC/NA/180; DSDP Site 607, c.c. 13; Late Pliocene;
female left valve, external and internal views; x 54. 17, GC/NA/155; DSDP Site 607, c.c. 5; Quaternary;
male left valve, external view; x45.

All figures are scanning electron micrographs.

PLATE 5

COLES et a/., Krithe

106

PALAEONTOLOGY, VOLUME 37

Remarks. This abundant, widely distributed subspecies is especially common in the Oligocene of
Hole 549A, but is rare or absent in Miocene to Quaternary sediments. Considerable size variation
occurs in this subspecies; there is a general but not consistent increase in size from the Middle
Eocene to the Oligocene. K. pernoides pernoides differs from its descendant, K. pernoides sinuosa
Ciampo, 1986 in having a regularly convex posterior margin, while in the latter it is more angular
with a marked posterior angle and posterior concavity. K. pernoides pernoides may have branching
ARPC and a ‘Y’-shaped anterior vestibulum, which are not present in K. pernoides sinuosa.

The original figures of Bornemann (1855) are poor and only the exterior of the carapace was
illustrated. The identification of the present specimens with K. pernoides (Bornemann) is based on
the records of subsequent authors who illustrated the internal features.

The known occurrences of this species are as follows: Upper Eocene of Denmark, eastern
Germany (Pietrzeniuk 1969) and the Ukraine (Schremeta 1969); Oligocene of Denmark,
Netherlands (Kuiper 1918), Hermsdorf near Berlin, Germany (Bornemann 1855), Belgium (Keij
1957), northern Germany (Uffenorde 1981); Miocene of northwestern Germany (Bassiouni 1962;
Uffenorde 1981); Upper Pliocene to Quaternary of the North Atlantic (Coles 1985) and late
Quaternary of the Atlantic west of the Iberian Portal at 2414 and 2421 m (Harpur 1985).

In summary, this species is common in deep-water Middle Eocene to Oligocene sediments in the
North Atlantic, but is rarer in the Miocene to Quaternary of the same area. In the Upper Eocene
to Miocene interval of Europe it occurs in shallower (outer shelf?) warmer waters from the evidence
of the associated fauna, and is frequently associated with such long-ranging and eurybathic species
as Henryhowella asperrima (Reuss).

Krithe pernoides sinuosa Ciampo, 1986

Plate 5, figures 13-17; Text-figure 5r-v

1962 Krithe pernoides (Bornemann); Ruggieri, p. 17, pi. 1, figs 12-13.

1976 Krithe pernoides (Bornemann); Breman; p. 55, pi. 3, fig. 28.

EXPLANATION OF PLATE 6

Figs 1-3. Krithe pernoides sinuosa Ciampo. 1, GC/NA/155; DSDP Site 607, c.c. 5; Quaternary; male left valve,
internal view; x45. 2-3, GC/NA/156; DSDP Site 607, c.c. 5; Quaternary; male right valve, external and
internal views; x45.

Figs 4-6. Krithe sp. cf. K. parvula Deltel. 4-5, GC/NA/927; DSDP Site S49, c.c. 8^1; middle Eocene; female
left valve, external and internal views; x95. 6, GC/NA/929; DSDP Site 549, c.c. 8-4; middle Eocene;
female right valve, external view; x93.

Figs 7-9. Krithe sp. 7. 7, GC/NA/900; DSDP Site 549A, c.c. 5-5; Late Miocene; female left valve, external
view; x 54. 8, GC/NA/903; DSDP Site 549A, c.c. 5-5; Late Miocene; male right valve, external view; x 47.
9, GC/NA/901 ; DSDP Site 549A, c.c. 5-5; Late Miocene; female right valve, internal view; x 55.

Figs 10-12. Krithe sp. 8. 10, GC/NA/269; DSDP Site 610E, c.c. 7; Late Miocene; female left valve, external
view; x 70. 11-12, GC/NA/270; DSDP Site 610E, c.c. 7; Late Miocene; female right valve, external and
internal views; x 70.

Figs 13-14. Krithe sp. 9. GC/NA/210; DSDP Site 607, c.c. 24; Early Pliocene; female left valve, external and
internal views; x 52.

Fig. 15. Krithe sp. 10. GC/NA/42; DSDP Site 606A, c.c. 19; Early Pliocene; female? right valve, external view;
x 52.

Fig. 16. Krithe sp. 11. GC/N A/206; DSDP Site 61 ID, c.c. 9; Early Pliocene; female left valve, external view;
x 66.

Figs 17-18. Krithe sp. 12. GC/NA/266; DSDP Site 609B, c.c. 29; Early Pliocene; female right valve, external
and internal views; x 60.

All figures are scanning electron micrographs.

PLATE 6

fflm

COLES et al. , Krithe

108

PALAEONTOLOGY, VOLUME 37

1977 Krithe sp. A Peypouquet, p. 104, fig. 34 (not sp. A24?).

1985 Krithe producta Brady; Guillaume et ah. pi. 106, fig. 5.

1986 Krithe sinuosa Ciampo, p. 87, pi. 17, figs 3, 5.

Material and distribution. One hundred and eight adult valves. Hole 549A (U. Miocene). Hole 550
(U. Miocene), Hole 558 (L.7-U. Miocene), Hole 563 (M.-U. Miocene), Holes 606, 606A, 607, 609, 610
(Pliocene-Quaternary), Hole 608 (U. Miocene-Quaternary), Holes 608A, 609B (U. Pliocene-Quaternary),
Holes 609A, 610C (Quaternary), Hole 610B (U. Pliocene), Hole 61 ID (L. Pliocene, Quaternary).

Dimensions {mm).

L

H

FLV GC/NA/180 607-13 U. Pliocene

0-77

0-43

FLV GC/NA/36 606-1 U. Pliocene

0-74

0-40

MLV GC/NA/155 607-5 Quaternary

0-94

0-44

MLV GC/NA/851 558 17-5 L. Miocene

0-80

0-36

FRV GC/NA/852 563 10-1 M. Miocene

0-69

0 35

FRV GC/NA/35 606A-5 Quaternary

0-75

0-40

MRV GC/NA/156 607-5 Quaternary

0-92

0-41

Stratigraphical range. Middle or possibly Lower Miocene to Quaternary (NN 1?, 5-7, 9-12, 14-21) in the
present study. The species also occurs in the Recent.

Diagnosis. A medium (female) to large (male), subrectangular and moderately inflated subspecies
of Krithe with strong sexual dimorphism. Females moderately high, males elongate to very elongate.
Both sexes with straight to slightly convex dorsal margin; posterior bluntly truncate with distinct
angle between shallow posterior concavity and steep postero-ventral slope. Anterior vestibulum
small to medium sized, ‘pocket ’-shaped. ADRPC type 3A. VRPC long, slightly curved; some are
branched.

Remarks. Krithe sinuosa was described from the Upper Miocene of Italy by Ciampo, but is here
considered to be a subspecies of K. pernoides Bornemann, evolving from the nominative subspecies
in the Early to Mid-Miocene. K. pernoides sinuosa seems to be an ecological subspecies of
K. pernoides , with the former dominating at abyssal depths, although the latter has been recorded
down to a PDWD of 3526 m at Site 608.

EXPLANATION OF PLATE 7

Figs 1-3. Krithe sp. 13. 1-2, GC/NA/260; DSDP Site 609B, c.c. 2; Quaternary; female left valve, external and
internal views; x 50. 3, GC/NA/227 ; DSDP Site 608, c.c. 5; Quaternary; female right valve, external view;
x 50.

Figs 4-6. Krithe sp. 14. 4, GC/NA/312; DSDP Site 61 ID, c.c. 12; Late Pliocene; female left valve, external
view; x 38. 5-6, GC/NA/170; DSDP Site 607, c.c. 10; Late Pliocene; female right valve, external and
internal views; x 39.

Figs 7-9. Krithe sp. 15. 7, GC/NA/293; DSDP Site 610, c.c. 5; Quaternary; male left valve, external view;
x 58. 8-9, GC/NA/294; DSDP Site 610, c.c. 5; Quaternary, male right valve, external and internal views;
x 59.

Figs 10-1 1 . Krithe sp. 16. GC/NA/262; DSDP Site 609B, c.c. 9; Quaternary; female? right valve, external and
internal views; x45.

Figs 12-13. Krithe sp. 17. GC/NA/52; DSDP Site 606A, c.c. 4; Quaternary; female? right valve, external and
internal views; x47.

Figs 14-15. Krithe sp. 18. GC/NA/166; DSDP Site 607, c.c. 6; Quaternary; male? right valve, external and
internal views; x49.

Figs 16-17. Krithe sp. 19. GC/NA/54; DSDP Site 606A, c.c. 2; Quaternary; female? right valve, external and
internal views; x46.

All figures are scanning electron micrographs.

PLATE 7

COLES et al Krithe

0008

110

PALAEONTOLOGY, VOLUME 37

text-fig. 6. For legend see opposite.

COLES ET A L. : OSTRACOD GENUS KRITHE

In the present material, K. pernoides sinuosa is common in Middle Miocene to Quaternary
sediments in the North Atlantic. A single MLV occurred in the Lower Miocene with an abnormally
large anterior vestibulum, but is otherwise very similar to typical K. pernoides sinuosa.

In addition to the occurrences cited above, the following are noted: Upper Miocene of Sicily
(Ruggieri 1962) and DSDP Site 403 on the Rockall Plateau (material of Ducasse and Peypouquet
1979 seen by GC), Pliocene of DSDP Site 613 off New Jersey (among the material of Cronin and
Compton-Gooding 1987 seen by GC), late Quaternary of the northeastern Atlantic in cores L4, L5,
P2, R3, R4, T2 and T3 between PDWD 1678 and 4566 m and lat. 43° and 68° N (among the
material of Porter 1984 seen by GC and RCW), late Quaternary of the North Atlantic west of the
Iberian Portal between 453 and 3700 m (Harpur 1985) and Quaternary of the western
Mediterranean, Gulf of Cadiz and Cape St Vincent between 795 and 2798 m (Elant 1985).

In the Recent it occurs in the Adriatic between 175 and 302 m (Breman 1976), the Atlantic around
the Iberian Portal and the western Mediterranean between 585 and 2798 m (Elant 1985), Atlantic
off northwest Africa between 997 and 1250 m (Barkham 1985), Porcupine Sea Bight between 1320
and 1942 m (Symonds pers. comm.), Atlantic off Florida between 185 and 739 m (material of
Cronin 1983 seen by GC), Gulf of Benin and Atlantic off the Ivory Coast (Peypouquet 1977), deep-
water Bay of Biscay (Guillaume et al. 1985), off south Norway between 300 and 630 m, Guff of
Mexico at 1500 m and North Atlantic at 5383 and 5440 m (GC personal observations).

In summary, K. pernoides sinuosa is common in the North Atlantic from the Middle Miocene to
the Recent with a maximum depth range of 185 to 5440 m and ranging from lat. 5° N to at least
68° N. It has, however, not been recorded from the Caribbean, or the Indian or Pacific oceans.

Krithe sp. cf. K. parvula Deltel, 1963
Plate 6, figures 4-6; Text-figure 5w-z

1964 Krithe parvula Deltel, p. 172, pi. 4, figs 86-89.

71985 Krithe parvula Deltel; Ducasse et al.. pi. 78, fig. 14.

Material and distribution. Sixty eight adult valves. Hole 549 (U. Palaeocene-M. Eocene) (NP 9-15).

text-fig. 6. a, Krithe sp. 2. GC/NA/891; DSDP Site 549, c.c. 19-3; late Palaeocene; adult right valve.
b-c. Krithe sp. 3. b, GC/NA/934; DSDP Site 550, c.c. 33-3; early Eocene; male? left valve, c, GC/NA/935;
DSDP Site 550, c.c. 30-3; early Eocene; female? right valve, d-e. Krithe sp. 5. d, GC/NA/932; DSDP Site
549, c.c. 4—1 ; middle Eocene; female right valve, e, GC/NA/932; DSDP Site 549, c.c. 4-1 ; middle Eocene;
male right valve, f-g. Krithe sp. 4. f, GC/NA/889; DSDP Site 550, c.c. 30-3; early Eocene; female left valve.
G, GC/NA/890; DSDP Site 550, c.c. 30-3; early Eocene; female right valve. H, Krithe sp. 6. GC/N A/887;
DSDP Site 549A, c.c. 9-5; late Oligocene; female left valve, i-k. Krithe sp. 7. 1, GC/NA/896; DSDP Site 549A,
c.c. 5-5; Late Miocene; female right valve, j, GC/NA/897; DSDP Site 549A, c.c. 5-5; Late Miocene; male
left valve, k, GC/N A/898; DSDP Site 549A, c.c. 5-5; Late Miocene; male right valve, l-m. Krithe sp. 8.
l, GC/NA/269; DSDP Site 610E, c.c. 7; Late Miocene; female left valve. M, GC/NA/270; DSDP Site 610E,
c.c. 7; Late Miocene; female right valve, n, Krithe sp. 9. GC/NA/210; DSDP Site 607, c.c. 24; Early Pliocene;
female left valve, o, Krithe sp. 10. GC/NA/42; DSDP Site 606A, c.c. 19; Early Pliocene; female? right valve,
p, Krithe sp. 11. GC/NA/206; DSDP Site 61 ID, c.c. 9; Early Pliocene; female left valve. Q, Krithe sp. 12.
GC/N A/266; DSDP Site 609B, c.c. 29; Early Pliocene; female right valve. R-s. Krithe sp. 13. R, GC/N A/260;
DSDP Site 609B, c.c. 2; Quaternary; female left valve, s, GC/NA/227; DSDP Site 608, c.c. 5; Quaternary;
female right valve, t-u. Krithe sp. 14. t, GC/NA/312; DSDP Site 61 ID, c.c. 12; Late Pliocene; female left
valve, u, GC/NA/170; DSDP Site 607, c.c. 10; Late Pliocene; female right valve, v-w. Krithe sp. 15.
v, GC/NA/293; DSDP Site 610, c.c. 5; Quaternary; male left valve, w, GC/NA/294; DSDP Site 610, c.c. 5;
Quaternary; male right valve, x, Krithe sp. 16. GC/NA/262; DSDP Site 609B, c.c. 9; Quaternary; female?
right valve, y, Krithe sp. 17. GC/NA/52; DSDP Site 606A, c.c. 4; Quaternary; female? right valve, z, Krithe
sp. 18. GC/NA/166; DSDP Site 607, c.c. 6; Quaternary; male? right valve, aa, Krithe sp. 19. GC/NA/54;
DSDP Site 606A, c.c. 2; Quaternary; female? right valve. All projectina drawings; x 50.

112

PALAEONTOLOGY, VOLUME 37

Dimensions (mm). L H

FLV GC/NA/925 549 11-1 U. Eocene

044

0-22

FLV GC/NA/926 549 11-1 U. Eocene

044

0-22

FLV GC/N A/927 549 8-4 M. Eocene

044

0-22

MLV GC/N A/932 549 16-3 U. Palaeocene

046

0-20

FRV GC/NA/928 549 8-4 M. Eocene

043

0-21

FRV GC/NA/929 549 8-4 M. Eocene

045

0-21

FRV GC/NA/930 549 11-1 L. Eocene

044

0-21

MRV GC/NA/931 549 11-1 L. Eocene

044

0-18

Diagnosis. A small, elongate, subrectangular (male and FLV) to elongate subovate (FRV) species
of Krithe with almost straight, subparallel dorsal and ventral margins. Anterior vestibulum
moderately large, ‘mushroom ’-shaped with a narrow neck and directed postero-ventrally. Inner
lamella wide, with sinuous line of concrescence and long RPC; ADRPC type 3A with AD 1, 2, 3
and 5 short to moderately long and elongate AD4. Sexual dimorphism slight; males relatively but
not absolutely longer than females.

Remarks. This small, distinctive species is common only in the Lower Eocene of Hole 549 and
disappears in the lower Middle Eocene (NP 15). K. parvula Deltel, 1964, described from the Middle
and Upper Eocene of Aquitaine, is very similar in shape and external features to the present
specimens. However, as Deltel did not illustrate the internal features, and her specimens are slightly
larger (L = 0-52 mm), the present material is only compared to K. parvula.

Nomina nuda species

Discussion. In addition to the seventeen species and subspecies detailed above, there are numerous
quite distinctive Krithe species present in the Cenozoic of the North Atlantic (Pis 6-7 ; Text-figs 5-6).
However, these are all too rare to be described formally; none of the species is represented by more
than seven adult specimens and many are known from only one or two individuals. The
stratigraphical ranges of nineteen nomina nuda species are given below, together with their
occurrence.

Species

Location

Stratigraphical position

Krithe sp. 1

DSDP Hole 549

U. Palaeocene, NP7-9

Krithe sp. 2

DSDP Hole 549

U. Palaeocene, NP 7

Krithe sp. 3

DSDP Hole 550

L. Eocene, NP 10

Krithe sp. 4

DSDP Hole 550

L. Eocene, NP 10

Krithe sp. 5

DSDP Hole 549

M. Eocene, NP 15-16

Krithe sp. 6

DSDP Hole 549A

U. Oligocene, NP 23-24

Krithe sp. 7

DSDP Hole 549A

U. Miocene, NN 9, 1 1

Krithe sp. 8

DSDP Hole 610

U. Miocene, NN 1 1

DSDP Hole 610E

U. Miocene, NN 10, 12

Krithe sp. 9

DSDP Hole 607

L. Pliocene, NN 13-14

Krithe sp. 10

DSDP Hole 606A

L. Pliocene, NN 14

Krithe sp. 1 1

DSDP Hole 607

L. Pliocene, NN 1 5

Krithe sp. 12

DSDP Hole 609B

L. Pliocene, NN 15

Krithe sp. 13

DSDP Hole 607

Quaternary, NN 19

DSDP Hole 608

L. Pliocene, Quaternary, NN 15, 18-19

DSDP Hole 609B

Quaternary, NN 19

DSDP Hole 610

U. Pliocene, NN 16

Krithe sp. 14

DSDP Hole 607

U. Pliocene, NN 16

DSDP Hole 61 ID

U. Pliocene, NN 18

Krithe sp. 15

DSDP Hole 609C

U. Pliocene, NN 16-17

DSDP Hole 609A

Quaternary, NN 20-21

DSDP Hole 610

Quaternary, NN 19

DSDP Hole 6 10C

Quaternary, NN 21

COLES ET AL.: OSTRACOD GENUS KRITHE

113

Species
Krithe sp. 16
Krithe sp. 17
Krithe sp. 18
Krithe sp. 19

Location

DSDP Hole 609B
DSDP Hole 606A
DSDP Hole 607
DSDP Hole 606A

Stratigraphical position
Quaternary, NN 19
Quaternary, NN 19
Quaternary, NN 19
Quaternary, NN 19-21

BIOSTRATIGRAPHY

Several authors have attempted to utilize Krithe in Cainozoic biostratigraphy, though usually in
conjunction with other genera. The work of van den Bold (1977) in the Caribbean region and the
South Atlantic is especially notable, while others have recognized some stratigraphically useful
species in the Mediterranean, e.g. Ascoli (1968), Sissingh (1972), and Ciampo (1980, 1986).
However, there has been relatively little work in the North Atlantic, mostly limited to the studies
of Ducasse and Peypouquet (1979) in the upper Cainozoic of the eastern North Atlantic and that
of Peypouquet (1977, 1979) with reference to Pleistocene glacial cycles. More recent biostrati-
graphical utilization of the genus, among other North Atlantic deep water taxa, are by Whatley and
Coles (1991) and Whatley (1993).

The neglect of Krithe in biostratigraphical studies from deep-water sediments, despite its high
abundance and diversity, is undoubtedly a consequence of the formidable problems of consistent
species discrimination within the genus. Nevertheless, the careful discrimination of species
throughout the Cainozoic of the North Atlantic in the present study has revealed the stratigraphical
utility of several common species. The stratigraphical ranges of the seventeen described species of
Krithe considered in this study are shown in Text-figure 7 to the level of the calcareous nannofossil

text-fig 7. Stratigraphical ranges of Krithe species in the Cenozoic of the North Atlantic, plotted against
chronostratigraphy and calcareous plankton biozones.

114

PALAEONTOLOGY, VOLUME 37

NP and NN zones of Martini (1971). Also included are the stratigraphical ranges of the nineteen
nomina mulct species listed above. This diagram clearly illustrates the importance of the genus as a
biostratigraphical marker in the Tertiary of the North Atlantic. The consistency of new appearances
in the succession bears impressive witness to the rapidity of evolution in this taxon in the deep sea.

Certain species are entirely {K. praemorkhoveni, K. gobanensis , K. regulare, K. cf. parvula ) or
mostly (K. crassicaudata , K. cf. hixvanneensis, K. pernoides pernoides) confined to the Palaeogene,
while others are confined to the Miocene to Quaternary interval (K. reversa , K. minima , K. aquilonia,
K. pernoides sinuosa ); the remainder span the boundary (K. aequabilis, K. dolichodeira ,
K. morkhoveni - all subspecies, and K. trinidadensis).

The most important first appearances for stratigraphical purposes- are of K. reversa (NN 6,
Middle Miocene), K. minima (NN 5, Middle Miocene), K. aquilonia (NN 1, Lower Miocene) and
K. pernoides sinuosa (NN 5, Middle Miocene), while the most notable last occurrences are
K. gobanensis (NP 23, Upper Oligocene), K. regulare (NP 25, Upper Oligocene), K. crassicaudata
(NP 23, Upper Oligocene) and K. praemorkhoveni (NP 16, Upper Eocene).

The data in Text-figure 7 and the stratigraphical range chart for all North Atlantic Cainozoic
benthonic Ostracoda given in Whatley and Coles (1991, fig. 4), indicate that this group in this
environment has considerable potential as biostratigraphical markers. This information is
augmented by that of Whatley (1993, table 1) who further extolled the virtues of benthonic
ostracods as biostratigraphical indices in deep-sea environments in the Pacific and Indian Oceans,
as well as the Atlantic.

DIVERSITY AND PALAEOCEANOGRAPH Y

The number of Krithe species present in each NP and NN zone in the North Atlantic may be used
to show diversity trends in the taxon through the Cainozoic. The results are shown in Text-

1 2 3 4 5 6 7 8 9101112 13| 14 15 16 1 7 1 18 19 20

21 22 | 23 24 25

1 2 3 4 5 | 6 7 8 | 9 10 11 12

13 14 15 16 17 18

19 20 21

PALAEOCENE

EOCENE

OLIGOCENE

MIOCENE

PLIOCENE

QUATER-

NARY

LOWER 1 UPPER

LOWER | MIDDLE | UPPER

LWR | UPPER

LOWER | MIDDLE UPPER

LOWER | UPPER

text-fig. 8. Recorded and cumulative Krithe species diversity in the Cainozoic of the North Atlantic, plotted
against chronostratigraphy and calcareous plankton biozones.

figure 8. Both recorded (which indicates the actual presence of a species in a given zone) and
cumulative (which discounts the effect of temporary ‘Lazarus’ absences within the overall
stratigraphical range of a species) have been calculated to enable comparison with Whatley and

COLES ET AL.\ OSTRACOD GENUS KRITHE

115

Coles (1991, table 1 and fig. 2) which show the diversity trends of the entire North Atlantic
Cainozoic deep-water ostracod fauna.

From Text-figure 8 it is evident that Krithe steadily increased in both recorded and cumulative
diversity from low levels in the Palaeocene to a peak in the late Oligocene (NP 23 zone).
Subsequently, recorded Krithe species diversity plummeted in the early Miocene only to rise in an
irregular series of peaks through the Neogene to a Cainozoic maximum in the Quaternary (NN 19
zone). However, the cumulative species diversity pattern shows no dramatic late Oligocene-early
Miocene diversity and records a much more gradual rise in diversity through the Neogene to the
same Quaternary maximum. It is this later pattern, by eliminating the effects of unequal sampling
and taphonomy, that probably most closely reflects the true diversity of Krithe in the study area,
although it certainly masks an environmental perturbation which was responsible for the temporary
disappearance of a considerable number of species at this time. A similar diversity decline for the
entire ostracod fauna of the Atlantic deep water is shown in Whatley and Coles (1991, fig. 2) who
suggest that this early and mid-Miocene phenomenon was probably brought about by more
sluggish circulation patterns in the North Atlantic consequent upon the closure of the Iberian Portal
and effective isolation of the Tethys (Whatley and Coles 1987). The global temperature decrease of
some 7-8 °C brought about by the expansion of the Antarctic Ice Cap (Stanley 1987) and the
spillage of cold North Polar bottom waters into the North Atlantic over the subsided
Greenland-Iceland-Faroes-Scotland Ridge, produced the equivalent of modern North Atlantic
Deep Water (Schnikter 1980) and in this medium this characteristic fauna has evolved.

The essential features of Krithe diversity thus show strong similarities with the diversity pattern
of the entire ostracode fauna demonstrated by Whatley and Coles (1991, fig. 2), namely the steep
diversity increase through the Palaeogene to a late Oligocene peak, considerable reduction in
diversity in the early Miocene and a stable to gradual diversity rise through the Miocene and
Pliocene to a peak in the Quaternary. The mid-Pliocene warming interlude seems to be reflected in
a diversity peak in Krithe followed by a late Pliocene decline. This is in turn succeeded by an early
Quaternary peak which is the highest diversity for the Cainozoic for this taxon.

ECOLOGY AND PALAEOECOLOG Y

All species of Krithe are blind and smooth-shelled with elongate subcylindrical carapaces being the
dominant shape-type. They are thus well adapted to an infaunal mode of life as burrowers in soft
sediment substrates, principally fine sands, muds, and deep-water G/ohigerina and nannofossil
oozes. Krithe is cosmopolitan, but is confined to fully marine waters with salinities greater than 35
parts per thousand. The very wide depth range of Krithe was first noted by Brady (1880) in the
Challenger report, who recorded species from depths of 15 to 1825 fathoms (27-3340 m).

In shallow waters, Krithe is mainly cryophilic and confined to low-energy, soft sediment
substrates, although a few warm-water species have been recorded (Whatley and Downing 1984).
It is in the deep sea, at bathyal (1000-2000 m) and abyssal (> 2000 m) depths, where it is both
abundant and diverse that Krithe is most significant. All records of deep-sea ostracod faunas
include Krithe , which is normally the dominant genus below 1000 m. Indeed, Krithe frequently
outnumbers all other ostracode species combined when the total number of valves is considered.
This is demonstrated in Table 2, which details the composition of the Cainozoic deep-water North
Atlantic ostracod fauna in terms of the percentage of individuals belonging to each genus for each
age. The ten genera which comprise an average of 5 per cent of the individuals in any one age are
shown, together with the percentage of total individuals in the other fifty-nine genera. These figures
are calculated from sixty-six samples containing 100 or more specimens, comprising twenty-six from
Hole 549, twenty-five from Hole 549A, one from Hole 550, eight from Hole 558 and six from Hole
563. Also included are thirty-three samples from the DSDP Leg 94 study of Whatley and Coles
(1987) which contained 100 or more specimens.

From Table 2, it can be seen that Krithe comprises a mean of some 50 per cent (range 45 to 55
per cent) of Eocene and Oligocene faunas, and becomes progressively more dominant during the

116

PALAEONTOLOGY, VOLUME 37

table 2. The percentage generic composition of the deep water North Atlantic fauna by age. The ten genera
which comprise 5 per cent or more of the individuals in any age are shown, with the percentage of total
individuals in the remaining fifty nine genera. (@ = genus absent; x = genus is present but comprises < 1 per
cent of the fauna in that age.) Ages are early (E), middle (M) and late (L) divisions of the Palaeocene (Pa),
Eocene (E), Oligocene (O), Miocene (M) and Pliocene (P) epochs, and of the Quaternary (Q) Period.

Age

No. studied samples

LPa

2

EE

13

ME

10

LE

11

EO

9

LO

10

EM

3

MM

3

LM

5

EP

10

LP

11

EQ

9

LQ

3

Argilloecia

7

5

15

11

9

10

4

1

1

4

4

3

5

Bairdia / Bairdoppilata

3

4

6

6

4

2

1

1

2

1

X

X

X

Cvtherella

2

14

7

9

6

3

X

5

1

1

X

X

X

Cytheropteron

@

1

1

X

1

2

1

1

1

2

4

5

4

Henryhowella

@

@

1

3

4

4

9

9

9

6

5

5

2

Krithe

69

55

46

49

51

48

45

52

60

64

68

67

67

Parakrithe

@

@

1

4

4

6

6

9

2

1

2

X

X

Poseidonamicus

@

@

@

@

1

5

4

3

11

8

6

6

10

T rachyleberidea

6

9

5

2

1

1

@

@

@

@

@

X

X

Xestoleberis

1

1

2

3

3

2

5

1

1

1

X

X

X

Others (59 genera)

12

11

16

13

16

17

25

18

12

12

11

14

12

Miocene, to comprise on average, some two-thirds (64 to 68 per cent) of Pliocene and Quaternary
faunas. This may be due to the greater palaeodepth of the Pliocene and Quaternary samples, relative
to those from the Eocene and Oligocene. Argilloecia or Cytherella are usually the second most
dominant genera in Palaeogene samples while Poseidonamicus is the next most abundant genus in
terms of individuals from the upper Miocene to the Quaternary.

A major palaeoecological application of Krithe in recent years has been based upon the supposed
relationship of anterior vestibular form with depth and the level of dissolved oxygen in sea water
(Peypouquet 1975, 1977). It is argued that in high levels of dissolved oxygen (> 04 ml), species of
Krithe with small vestibula and wide inner lamellae would predominate while in waters where
oxygen levels were low ( < 04 ml), vestibula would be large with narrow inner lamellae. It is further
suggested that as a consequence of this relationship, past oxygen levels could be determined from
the size of the anterior vestibulum and inner lamella, leading to the reconstruction of past oceanic
oxygen levels. An attempt has been made to apply this to Cainozoic palaeoenvironments in the
Atlantic (Peypouquet 1977, 1979) and to the study of Cretaceous-Tertiary boundary events in
Tunisia (Peypouquet 1983).

In an attempt to investigate the validity of this hypothesis, the present-day water depth range of
eleven species or subspecies which occur in the Quaternary and Recent of the North Atlantic have
been calculated (see the taxonomic section for further details), and compared with their size and
anterior vestibulum and ADRPC type, shown in Table 3.

Table 3 shows that those species with the broadest depth distribution are K. dolichodeira and
K. pernoides sinuosa with large and small anterior vestibula respectively. Although species with
small vestibula predominate at abyssal depths, the deepest known Krithe species in this study is
K. reversa at 5726 m, which has a large ‘pocket '-shaped anterior vestibulum. Therefore, the data
from the present study cannot support the correlation of vestibular size with depth and,
consequently, oxygen levels as proposed by Peypouquet (1977, 1979).

In addition, there is no obvious link between ADRPC type and water depth, with all three main
types being present at abyssal depths. However, most species with ADRPC type 3A except the
K. pernoides (i.e. the K. trinidadensis group) are confined to depths below 1000 m. These findings
confirm those of Whatley and Zhao (1993) and Zhao and Whatley (1993) that there is no
relationship between the size and shape of the anterior vestibulum, and depth or oxygen
concentration. These authors also were able to discount the assertion by Peypouquet (1977, 1979)

COLES ET AL.: OSTRACOD GENUS KRITHE

117

table 3. The present-day water depth range, anterior vestibulum type and antero-dorsal radial pore canal type,
of the eleven Krithe species and subspecies present in the Quaternary and Recent of the North Atlantic.

Species/subspecies

PDWD range

Vestibule type

ADRPC

type

K. reversa

803-5726 m*

Large ‘pocket’

IB

K. aequabilis

1 200-4000 m

Large ‘pocket’

2B

K. dolichodeira

200-5440 m

Large ‘mushroom’

2B

K. minima

739-4566 m*

Small crescentic

2C

K. mork. morkhoveni

739-5440 m*

Small ‘mushroom’

3A

K. mork. lamellalata

2421-3422 m

Small ‘mush.’/‘Y’

3A

K. mork. ayressi

1 500-3884 m

Large ‘mushroom’

3A

K. trinidadensis

1 080-5440 m

Small ‘pock. ’/‘ Y'

3A

K. aquilonia

938-4566 m

Small ‘mush. ’/‘Y’

3A

K. per. pernoides

2414-3526 m

Small ‘pocket’

3 A

K. per. sinuosa

175-5440 m

Small ‘pocket’

3A

* Almost all records are from water depths in excess of 1000 m.

that the overall size of species of Krithe increases with depth; this is confirmed by the size
distribution with depth of the taxa considered in the present paper.

Acknowledgements. The authors thank their colleagues (Michael Ayress, Simon Barkham, Howard Davies,
Sian Downing, Chris Harlow, Will Harpur and Christine Porter) at Aberystwyth for access to their research
material and numerous valuable discussions. In particular, Chris Harlow is acknowledged for his part in the
original formulation of the antero-dorsal radial core canal classificatory scheme for Krithe. Ian Gully and
David Griffiths are thanked for their skills in draughtmanship and photography respectively. Graham Coles
wishes to acknowledge the NERC studentship which made this study possible. Officers of DSDP/ODP are
thanked for sending the various samples upon which much of this work is based.

REFERENCES

ascoli, p. 1964. Preliminary ecological study on Ostracoda from bottom cores of the Adriatic Sea.
Pubblicazioni della Stazione Zoologica di Napoli , 33, 213-246.

— 1968. Preliminary report on the Ostracoda of the type Tortonian. Giornale di Geologia , 35 (2), 31-54.
ayress, M. a. 1988. Late Pliocene to Quaternary Deep-Sea Ostracoda from the eastern Indian and southwestern

Pacific Oceans. Unpublished Ph D. thesis. University of Wales, Aberystwyth.
baird, w. 1850. The natural history of the British Entomostraca. Ray Society of London, London, 364 pp.
barkham, s. t 1985. Recent Ostracoda from off the N.W. African coast. Unpublished M.Sc. thesis. University
of Wales, Aberystwyth.

bassiouni, M. a. a. 1962. Ostracoden aus dem Mittelmiozan in N.W. Deutschland. Romeriana , 3, 1-123.
benson, r. h. and peypouquet, j.-p. 1983. The upper and mid-bathyal Cenozoic ostracode faunas of the Rio
Grande Rise found on Leg 72 Deep Sea Drilling Project. Initial Reports of the Deep Sea Drilling Project, 72,
805-818.

bold, w. a. van den 1946. Contribution to the study of Ostracoda with special reference to the Tertiary and
Cretaceous microfauna of the Caribbean region. Dissertation, Utrecht University. Debussy, Amsterdam.
— 1958. Ostracoda of the Brasso Formation of Trinidad. Micropaleontology , 4, 391-418.

1960. Eocene and Oligocene Ostracoda of Trinidad. Micropaleontology , 6, 145-196.

— 1966. Ostracoda of the Pozon Section, Falcon, Venezuela. Journal of Paleontology, 40, 175-185.

— 1968. Ostracoda of the Yague group (Neogene of the northern Dominican Republic). Bulletin of
American Paleontology, 54, 239-266.

1969. Messinella, a new genus of Ostracoda in the Caribbean Cenozoic. Micropaleontology, 15, 397^100.
1977. Cenozoic marine Ostracoda of the South Atlantic. 495-519. In swain, f. m. (ed.). Stratigraphic
micropaleontology of Atlantic Basin and borderlands. Elsevier, Amsterdam.

PALAEONTOLOGY, VOLUME 37

bold, w. a. ven den 1981. Distribution of Ostracoda in the Neogene of central Haiti. Bulletin of American
Paleontology , 79. 7-136.

bornemann, j. G. 1855. Die mikroskopische Fauna des Septarienthones von Hermsdorf bei Berlin. Zeitschrift
der Deutschen Geologischen Gesellschaft , 7, 1-167.

brady, G. s., crosskey, h. w. and robertson, d. 1874. The Post-Tertiary Entomostraca. Monograph of the
Palaeontographical Society , 28 (127), i-v+ 1-232.

breman, e. 1975. Ostracoda in a bottom core from the deep south-eastern basin of the Adriatic Sea. Koninklijke
Nederlandse Akademie van Wettenschappen , Series B, 78, 198-218.

— 1976. The distribution of ostracodes in the bottom sediments of the Adriatic Sea. Vrije Universiteit
Amsterdam, 165 pp.

breman, G. s. 1880. Report on the Ostracoda dredged by H.M.S. Challenger during the years 1873-1876.

Report of the Scientific Results, Voyage of H.M.S. Challenger, Zoology, 1, 1-184.
bubikyan, s. a. 1958. [Ostracods of the Paleogene deposits of the Erivan Basin.] Izvestiya Akademii Nuak
armyanskoi, Erevan, SSR (series Geology-Geography), 11, 3-16. [In Russian.]
ciampo, G. 1980. Ostracodi Miocenici (Tortoniano-Messimano) della regione di Ragusa (Sicilia). Bolletino
della Societd Paleontologica Italiana, 19, 5-20.

— 1981. Ostracoda fossili (Oligocene superiore-Serravalliano) del Monte Cammarata (Sicilia centro-
occidentale) e del Ragusano (Sicilia sud-orientale). Bolletino della Societd Paleontologica Italiana, 20, 53-72.

1986. Ostracodi del limite Tortoniano/Messiniano in alcune sezioni italiane. Bolletino della Societd
Paleontologica Italiana, 24, 29-110.

coles, G. p. 1985. Upper Miocene, Pliocene and Quaternary Krithe and Parakrithe from DSDP Leg 94,
N. Atlantic. Unpublished M.Sc. thesis. University of Wales, Aberystwyth.

— and whatley, r. c. 1989. New Palaeocene to Miocene genera and species of Ostracoda from DSDP Sites
in the North Atlantic. Revista Espanola de Micropaleontologia, 21, 81-124.

coxill, d. j. 1985. Deep water Recent Ostracoda from the South Atlantic around South Georgia, South
Sandwich Islands and the Antarctic Peninsula and the littoral ostracods of the Falkland Islands.
Unpublished M.Sc. thesis. University of Wales, Aberystwyth.
cronin, T. m. 1983. Bathyal ostracodes from the Florida-Hatteras Slope, the Straits of Florida and the Blake
Plateau. Marine Micropaleontology, 8, 89-119.

— and compton-gooding, e. e. 1987. Cenozoic Ostracoda from DSDP Leg 95 off New Jersey (Sites 612 and
613). Initial Reports of the Deep Sea Drilling Project, 95, 439 — 45 1 .

dainty, e. j. 1984. Pleistocene Ostracoda from four DSDP Sites of the South-West Pacific. Unpublished M.Sc.
thesis. University of Wales, Aberystwyth.

davies, h. c. 1981. The areal and depth distribution of N.E. Atlantic Ostracods. Unpublished M.Sc. thesis,
University of Wales, Aberystwyth.

deltel, b. 1964. Nouveaux ostracodes de l’Eocene et de l’Ohgocene de l'Aquitaine meridionale. Actes de la
Societe Linneenne de Bordeaux, 100, 127-221.

dieci, g. 1967. Riesame di alcune specie di ostracodi Tortoniani dell’Appennino settentrionale. Bolletino della
Societd Paleontologica Italiana, 6, 3-17.

— and russo, a. 1964. Ostracodi Tortoniani delPAppennino settentrionale (Tortona, Montegibbio,
Castelvetro). Bolletino della Societd Paleontologica Italiana, 3, 38-88.

dingle, r. v., lord, a. r. and boomer, i. d. 1990. Deep-water Quaternary Ostracoda from the continental
margin off south-western Africa (SE Atlantic Ocean). Annals of the South African Museum, 99, 245-366.
downing, s. e. 1985. The taxonomy, palaeoecology, biostratigraphy and evolution of Pliocene Ostracoda from
the W. Pacific. Unpublished Ph.D. thesis, University of Wales, Aberystwyth.
ducasse, o., guernet, c. and tambareau, y. 1985. Paleogene. 257-311. In oertli, h. j. (ed.). Atlas des
ostracodes de France. Bulletin du Centre de Recherches Exploration et Production, Elf- Aquitaine, 9, 1-396.

— and peypouquet, j.-p. 1979. Cenozoic ostracodes: their importance for bathymetry, hydrology and
biogeography. 343-363. In montadert, l. and Roberts, d. g. (eds). Initial Reports of the Deep Sea Drilling
Project, 48. US Government Printing Office, Washington.

elant, m.-o. 1985. Les ostracodes et les relations hydrologiques Atlantique-Mediterranee du dernier glaciaire
a l’actuel. Thesis, Universite de Bordeaux no. 2065.

Guillaume, m.-c., peypouquet, j.-p. and tetart, J. 1985. Quaternaire et Actuel. 337-377. In oertli, h. j. (ed.).
Atlas des Ostracodes de France. Bulletin du Centre de Recherches Exploration et Production, Elf- Aquitaine,
9, 1-396.

hanai, t. 1959. Studies on the Ostracoda of Japan. Part 4. Family Cytherideidae Sars, 1925. Journal of the
Faculty of Science, Tokyo University, 2, 291-308.

COLES ET AL.: OSTRACOD GENUS KRITHE

119

harpur, w. k. 1985. Late Quaternary deep-sea Ostracoda from the extra-Iberian Portal region. Unpublished
M.Sc. thesis. University of Wales, Aberystwyth.

howe, h. v. and law, j. 1936. Louisiana Vicksburg Oligocene Ostracoda. Louisiana Geological Survey Bulletin ,
7, 1-96.

hulings, n. c. 1967. Marine Ostracoda from the Western North Atlantic Ocean between Cape Hatteras, North
Carolina, and Jupiter Inlet, Florida. Bulletin of Marine Science, 17, 629-659.
jones, T. R. and kirkby, j. w. 1989. On Carboniferous Ostracoda from Ireland. Scientific Transactions of the
Royal Dublin Society, Series 2, 6, 173-200.

keij, a. j. 1957. Eocene and Oligocene Ostracoda of Belgium. Memoires de I Inst it ut Royal des Sciences
Naturelles de Belgique, 136, 1-210.

kempf, e. k. 1988. Index and bibliography of marine Ostracoda. 1, Index A. Geologisches Institut der
Universitdt zu Koln Sonderverojfentlichungen, 50, 1-762.
kuiper, w. N. 1918. Oligocane und Miocane Ostracoden aus den Niederlanden. Proefschrift Gebroeders
Hoitsema, Groningen, 1-91.

latreille, p. a. 1806. Genera crustaceorum et insectorum. Koenig, Paris.

leroy, d. o. and Levinson, s. a. 1974. A deep-water Pleistocene micro-fossil assemblage from a well in the
northern Gulf of Mexico. Micropaleontology , 20, 1-37.
martini, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. Proceedings of the
2nd Planktonic Conference, Rome, 2, 739-785.

McKenzie, k. g., majoran, s., emami, v. and reyment, r. a. 1989. The Krithe problem - first test of
Peypouquet’s hypothesis, with a redescription of Krithe praetexta praetexta (Crustacea, Ostracoda).
Palaeogeography, Palaeoclimatology, Palaeoecology, 74, 343-354.
morkhoven, f. p. c. m. van 1972. Bathymetry of Recent marine Ostracoda in the North-West Gulf of Mexico.

Transactions of the Gulf-Coast Association of Geological Societies, 22, 241-252.
muller, g. w. 1 894. Die Ostracoden des Golfes von Neapel und der angrezenden Meeres-Abschnitte. Fauna
und Flora des Golfes von Neapel, 21, 1-404.

oertli, h. j. 1961. Ostracodes du Langhien-type. Rivista Italiana di Paleontologia, 67, 17^44.
peypouquet, j.-p. 1975. Les variations des caracteres morphologiques internes chez les ostracodes des genres
Krithe et Parakrithe : relation possible avec le teneur en O,, dissous dans l’eau. Bulletin de /’ Institut de Geologie
du Bassin Aquitaine, 17, 81-88.

— 1977. Les ostracodes et la connaissance des paleomilieux profonds. Application au Cenozoi'que de
P Atlantique Nord-Oriental, Universite de Bordeaux, 433 pp.

1979. Ostracodes et paleoenvironnements. Methodologie et application aux domaines profonds du

Cenozoi'que. Bulletin , Bureau du Recherches Geologiques et Minieres, 2eme serie, 4, 3-79.

— 1983. Krithe and Parakrithe in the Kef Section (Northeast Tunisia) around the Cretaceous-Tertiary
boundary: palaeoecological implications. 510-519. In maddocks, r. f. (ed.). Applications of Ostracoda.
Department of Geosciences, University of Houston, 677 pp.

pietrzeniuk, E. 1969. Taxonomische und biostratigraphische Untersuchungen an Ostracoden des Eozan 5 im
Norden der Deutschen Demokratischen Republik. Paldontologische Abhandlungen, 4, 1-162.
pokorny, v. 1980. The genus Krithe (Ostracoda, Crustacea) in Palaeogene deep-sea deposits in the Zdanice
Unit, Moravia, Czechoslovakia. Casopsis pro Mineralogii a Geologii, 25, 338-349.
porter, c. 1984. Late Quaternary Ostracoda from the N.E. Atlantic. Unpublished M.Sc. thesis, University of
Wales, Aberystwyth.

purl h. s. and hulings, n. c. 1976. Designation of lectotypes of some ostracodes from the Challenger
Expedition. Bulletin of the British Museum ( Natural History), Zoological Series, 29, 251-314.
ruggieri, G. 1962. Gli ostracodi marini del Tortoniano (Miocene medi superiore) di Enna, nella Sicilia
centrale. Paleontologia italiana Memorie, 2, 1-68.

1974. Revisione della ostracofauna marina Quaternaria di Imola (Bologna). Revista Espahola de
Micropaleontologia, 6, 4 1 9 — 446.

russo, a. 1968. Ostracodi Tortoniani di Montebaranzone (Appennino settentrionale Modenese). Bolletino
della Societd Paleontologica Italiana , 7, 6-56.

sars, g. o. 1866. Oversigt af Norges marine ostracodes. Forhandlinger i Videnskabsselskabet i Kristiania, 7,
1-130.

schnitker, d. 1980. Global palaeoceanography and its deep-water linkage to the Antarctic glaciation. Earth
Sciences Reviews, 16, 1-20.

schremeta, v. 1969. The Paleogene ostracods of the Ukraine. Isdatel Lvov University, 274 pp.

120

PALAEONTOLOGY, VOLUME 37

seguenza, G. 1880. Le Formazioni Terziarie nella provincia di Reggio (Calabria). Memorie della R. Accademia
del Lincei. Classe di scienze fisiche e naturali, 6, 1-443.
shakin, a. 1991. Cenomanian/Turonian ostracods from Gelel Nezzarat, southwestern Sinai, Egypt, with
observations on <S13C values and the Cenomanian/Turonian boundary. Journal of Micropalaeontology, 10,
133-149.

sissingh, w. 1972. Late Cenozoic Ostracoda of the south Aegean Island Arc. Utrecht Micropaleontological
Bulletins , 6, 1-187.

smith, p. d. 1983. Quaternary deep sea Ostracoda from the Southwest Pacific. Unpublished M.Sc. thesis.
University of Wales, Aberystwyth.

steineck., p. L. 1981 . Upper Eocene to Middle Miocene ostracode faunas and paleo-oceanography of the North
Coastal Belt, Jamaica, West Indies. Marine Micropaleontology, 6, 339-366.

— breen, m., nevins, n. and o’hara, p. 1984. Middle Eocene and Oligocene deep-sea Ostracoda from the
Oceanic Formation, Barbados. Journal of Paleontology , 58, 1463-1496.

— dehler, d., hoose, e. and McCalla, d. 1988. Oligocene to Quaternary ostracodes of the Central
Equatorial Pacific (Leg 85, DSDP-IOPD). 597-617. In hanai, t., ikeya, n. and ishizaki, k. (eds).
Evolutionary biology of Ostracoda. Its fundamentals and applications. Kodansha-Elsevier, Amsterdam,
1361 pp.

uffenorde, h. 1981 . Ostracoden aus dem Oberoligozan und Miozan des unteren Elbe-gebeites (Niedersachsen
und Hamburg, NW Deutsches Tertiarbecken). Paldontologisches Abhandlungen, A , 172, 103-198.
whatley, R. c. 1970. Scottish Callovian and Oxfordian Ostracoda. Bulletin of the British Museum (Natural
History ), Geological Series, 19, 300-358.

— 1993. Ostracoda as biostratigraphical indices in Cainozoic deep-sea sequences. High Resolution
Stratigraphy, 70, 155-167.

— and coles, G. p. 1987. The Late Miocene to Quaternary. Ostracoda of Leg 94, Deep Sea Drilling Project.
Revista Espahola de Micropaleontologi'a, 19, 33-97.

— 1990. Global change and the biostratigraphy of North Atlantic Cainozoic deep water Ostracoda.
Journal of Micropalaeontology, 9, 119-132.

— and downing, s. e. 1983. Middle Miocene Ostracoda from Victoria, Australia. Journal of Micro-
palaeontology, 2, 83-104.

— and zhao quanhong 1993. The Krithe problem; a case history of the distribution of Krithe and
Parakrithe (Crustacea, Ostracoda) in the South China Sea. Palaeogeography, Palaeoclimatology,
Palaeoecology, 103, 281-297.

yassini, i. 1969. Ecologie des associations d’ostracodes du Bassin d’Arcachon et du littoral Atlantique.
Application a l’interpretation de quelques populations du Tertiaire Aquitaine. Bulletin de I Institut de
Geo/ogie du Bassin d Aquitaine, 7, 1-288.

zhao quanhong and whatley, r. c. 1993. [The ostracod genera Krithe and Parakrithe in the South China Sea:
implications for palaeoceanography.] 141-156. In ye zhizheng and wang pinxian (eds). Contributions to
late Quaternary paleoceanography of the South China Sea. Qingdao Ocean University Press, Beijing, 329 pp.
[In Chinese with English abstract.]

G. P. COLES
R. C. WHATLEY
A. MOGUILEVSKY

Micropalaeontology Research Group
Institute of Earth Studies

Typescript received 16 January 1993 University of Wales

Revised typescript received 1 July 1993 Aberystwyth, Dyfed SY23 3DB, UK

Editorial note. It is the policy in Palaeontology to use the International Union of Geological Sciences
chronostratigraphical nomenclature. The authors of this paper are opposed to the spelling ‘Cenozoic’ given
in the IUGS scheme, and prefer the etymologically more correct ‘Cainozoic’. In view of the strength of feeling
they expressed on this matter, the Editorial Board has used ‘Cainozoic’ in this paper.

SACCOCOM A: A BENTHIC CRINOID FROM THE
JURASSIC SOLN HO FEN LIMESTON E, GERM AN Y

by CLARE V. MILSOM

Abstract. Saccocoma is the most numerous macrofossil in the Solnhofen Limestone but its study has been
relatively neglected. Functional interpretations of the morphology and mode of life have been based on Jaekel’s
definitive paper (1892) in which the lifestyle of Saccocoma was considered to be pelagic. However, new
morphological interpretations suggest that Saccocoma may have been benthic. This conflicts with the
conventional interpretation of the Solnhofen environments which proposes that adverse conditions precluded
colonization by benthos.

Saccocoma was first illustrated by Bajer in 1730 (fide Bather 1924), and referred to by Goldfuss
(1829), who assigned it to the order Comatulida. It was described and named by Agassiz in 1835.
Quenstedt (1852) placed the saccocomids in the Asteroida. The Saccocomidae was established by
d’Orbigny (1852), but he did not place the family in a phylogenetic context. Zittel (1879) redescribed
the material of Goldfuss (1829), emphasizing the similarities between the saccocomids and crinoids
rather than ophiuroids. The most detailed description of Saccocoma was given in Jaekel (1892). This
exhaustive work has formed the basis of all current theories of the functional morphology and mode
of life of Saccocoma. Two main species were described: S. tenella and the juvenile equivalent
S. pectinata (Jaekel 1892). Walther (1904) included Saccocoma in his review of the fauna of the
Solnhofen Plattenkalk and described a new species, S. schwertschlageri. Recent work has involved
the identification and revision of the species belonging to Saccocoma. Key specific descriptions were
given by Sieverts-Doreck (1955), Verniory (1960 [5. tenella ], 1961 [5. quenstedti ], 1962), Hess (1972
[S', schattenbergi and S. subornata ]) and Manni and Nicosia (1984 [S. vernioryi]). Manni and Nicosia
(1985) regarded S. schwertschlageri as a junior synonym of S. tenella.

Saccocoma is the most numerous macrofossil in the Solnhofen lithographic limestones. This fossil
Lagerstatte (Seilacher et al. 1985), famous for the preservation of medusae and Archaeopteryx with
feathers, was deposited in the late Jurassic. The most fossiliferous unit of the Lagerstatte is the
Upper Solnhofen Plattenkalk which was formed in the Early Tithonian (ti2b) in the area extending
from Langenaltheim to Kelheim. S. tenella (including its juvenile stage S. pectinata) are most
abundant in the Eichstatt quarries where slabs covered with the crinoid occur. The abundance of
Saccocoma decreases in a south-westerly direction towards the Solnhofen quarries where specimens
are relatively rare.

Saccocoma tenella is also common in the Upper Kimmeridgian of Kimmeridge Bay (Arkell 1947)
where, unlike in the Solnhofen Limestone, preservation is poor. Only isolated ossicles are found and
these are rarely complete. S. tenella also occurs in the Kimmeridgian at Talloires near Lake Annecy
in the Haute-Savoie district of France (Verniory I960, 1962). S. vernioryi is reported from the
Tithonian of Marche, Italy (Manni and Nicosia 1985). S. tenella and S. quenstedti occur in the
Upper Oxfordian of Sisteron, Provence, France, the Malm 6 (= Middle Kimmeridgian (ki2)) of
Wurtemburg, Germany (Verniory 1961, 1962), and the late Jurassic of the North American Atlantic
seaboard (Hess 1972). S. schattenbergi comes from the Middle to Upper Oxfordian at Faucigny,
Haute-Savoie, France (Verniory 1962), and the late Jurassic of the American Atlantic seaboard
(Hess 1972). S. subornata occurs in the ‘Weissjura Gamma’ (Lower Kimmeridgian) of Southern
Germany and the late Jurassic of the North American Atlantic seaboard (Hess 1972).

| Palaeontology, Vol. 37, Part 1, 1994. pp. 1 211 29.|

© The Palaeontological Association

122

PALAEONTOLOGY, VOLUME 37

GENERAL MORPHOLOGY

The calyx of Saccocoma is bowl shaped with a convex base. It is composed of one main circlet of
five large radial plates, five very small basals and a minute centrale. These enclose a large, undivided
cup-shaped body cavity. The radial plates have a convex surface and taper aborally. They are
connected at their lateral margins by a prominent zig-zag suture. The articular facet is medially
placed at the position of the oral margin. The tegmen is not preserved.

The radials form the main part of the calyx. Their inner surface is smooth but the exterior is
covered with a complicated reticulate sculpture (Text-fig. 1). The network of anastomosing ridges

text-fig. 1. A, reticulate ornament on the radial plates of Saccocoma tenella , Kimmeridge Clay, Dorset; x 125.

b, detail of anastomosing ridges; x 1250.

is more irregular in the centre of the plate, becoming more organized towards the plate margins
where the ridges are broadly parallel and perpendicular to plate sutures. The lateral edges are
denticulate and the suture between two radials is formed by the interdigitation of the ridges. The
projecting ridges are inserted between the neighbouring ridges, thus forming a zig-zag suture. The
plates are extremely thin and feather out towards the lateral margins so that only the anastomosing
network of the outer ridges remains at the plate edges.

The articular facet is positioned at the centre of the ventral margin of the radial plate. It is small,
horizontal and lacks any distinct type of articulation.

The arms are uniserial. They dichotomize on the second brachial and distally branch
holotomously. The arms can therefore be divided into three parts (Text-fig. 2); the proximal part,
the secundiarm, and the holotomously branched region. The proximal part is an atomous section of
the arm consisting of two brachials only, the first primibrach IBr, and the primaxil IBr.,.

Primibrach IB^ is a simple, small, cylindrical ossicle which tapers distally. The ventral surface is
characterized by two distal grooves which diverge distally. Primaxil IBr2 is ornamented with large
wing-like flanges which are attached to the lateral margins of the brachial. These wings are concave
and smooth dorsally. The secundiarm consists of around seventeen unbranched secundibrachials
showing three morphological styles. IIBrj and II Br;J correspond to hypozygal brachials of syzygial

MILSOM: CRINOID SACCOCOMA

123

□ PROXIMAL BRACHIALS

SECUNDIBRACHIALS

HOLOTOMOUSLY BRANCHED AREA

text-fig. 2. Division of the arm of Saccocoma into three parts. Redrawn from Jaekel (1892).

pairs which in other crinoids lack pinnules. Brachials IIBr1 3 are small and unornamented; in
contrast, IIBr2 4_7 are ornamented with the paired lateral wing-like expansions seen on the IBr,
ossicle. Paired processes also occur on IIBr8_17 but they are attached to the lateral margins of the
aboral brachial surface and are parallel to each other and rectangular in outline. In the
holotomously branched part of the arm ramules are added regularly on alternate sides of every
other brachial. Ramule length decreases distally. The brachials are almost identical to the distal
secundibrachials except that they are smaller and the lateral processes are reduced.

PREVIOUS INTERPRETATIONS OF FUNCTIONAL MORPHOLOGY AND

MODE OF LIFE

The most detailed functional interpretation of the morphology of Saccocoma was given by Jaekel
(1892); he highlighted the delicate nature of the ossicles, particularly the radials, where weight had
been reduced to such an extent that only the anastomosing ridges exist at the radial margins. Jaekel
interpreted these ridges as structural supports which were necessary in order to compensate for the
loss of strength resulting from extreme weight reduction. As the network of ridges occurs on the
most distal brachials, he concluded that weight reduction occurred in all ossicles. This weight
reduction combined with the absence of a mode of attachment formed the foundation of the pelagic
hypothesis.

Jaekel (1892) interpreted the lateral flanges or ‘wings’ which occur on the lBr2 and IIBr, 4_7
brachials as supports for a membrane which would expand across this area, thus connecting the
lower parts of all arms, as in an octopus. Therefore the function of these ‘wings’ would be to
strengthen parts of the connecting skin. He suggested that the size difference between the adradial
‘wings’ and the larger abradial ‘wings’ prevented interference between wings of the same arm
during locomotion. Interpolated with these winged brachials are the wingless II Br4 and IIBr3
ossicles. According to Jaekel these ossicles aided mobility, and without them the wings of adjacent
brachials would interfere.

The ventrally directed vertical projections which occur from IIBr8 onwards Jaekel (1892) also
interpreted as soft tissue supports, but in this case rather than connecting the arms radially, the

124

PALAEONTOLOGY, VOLUME 37

tissue extended along the length of the arm forming a continuous channel within which the tube feet
could operate. Without these extensions Jaekel suggested that the activity of the tube feet would
have been limited due to the narrowness of the arms. Jaekel cited these structures as further evidence
of skeletal lightening.

The calyx is too small to enclose an efficient buoyancy aid, such as a gas bubble or fatty deposit;
however, Jaekel (1892) predicted that a buoyancy aid would be unnecessary or only a small aid
would be required to offset the minimal weight of the skeleton. Thus he concluded that Saccocoma
was planktonic, passively driven along with the current, in a similar manner to a jellyfish, with
mouth down and arms pendant.

Further to Jaekel’s (1892) interpretation, Seilacher et al. (1985) suggested that the necrolytic
deformation of the arms of the crinoids in the Solnhofen Limestone may be indicative of their
flexibility patterns during life. In Saccocoma the arms are distally coiled, in contrast to the
comatulids in which the proximal part is coiled but the arm tips are conspicuously straight. Thus
Seilacher et al. concluded that whilst the proximal arm section controls swimming in comatulids,
the distal portion was responsible in saccocomids.

Although these functional interpretations are very feasible, they do not represent the only
possible scenario. Certain morphological factors have been neglected whilst others require further
consideration.

ADDITIONAL EVIDENCE OF MODE OF LIFE FROM FUNCTIONAL

MORPHOLOGY

Specific gravity

Saccocoma would have been able to stay suspended in mid-water if its density was equal to that of
sea water. The crinoid endoskeleton is composed of high-magnesium calcite with a density
significantly greater than sea water. As the soft tissue is also denser than sea water it will not
compensate for endoskeleton density. Therefore Saccocoma must either have been continuously
active in order to generate lift economically from its forward motion, or have had a special
buoyancy aid. Jaekel (1892) and Seilacher et al. (1985) suggested a medusoid mode of life for
Saccocoma , thus implying the presence of a fluid with a low specific gravity contained within the
body tissue. A large volume of low density fluid would be required to offset the excess skeletal and
tissue density. In extant medusae, buoyancy fluids only give L5 mg lift in sea water for every cubic
centimetre of fluid (Denton 1974). This is very little lift and it is only because these animals are
unskeletonized that the fluids make the animal buoyant. Even so, jellyfish must swim actively to
avoid sinking through the water column. The presence of a fossilized skeleton and the lack of
evidence for a large amount of low density material argues against a pelagic mode of life for
Saccocoma.

Orientation

Jaekel (1892) suggested that Saccocoma floated mouth downwards in the water column. Such an
orientation would severely hinder the efficiency with which the crinoid fed. Gislen (cited in Peck
1955) proposed that roveacrinids would have difficulty catching sinking particles if they floated with
the mouth downwards. Furthermore, the calcareous dorsal parts would be appreciably heavier than
the fleshy ventral parts and would favour a mouth-upwards orientation.

Presence of flanges on IBr2 and IIBr2 4_7

Analysis of disarticulated specimens of Saccocoma tenella from the Upper Kimmeridge Clay of
Kimmeridge Bay has revealed that these flanged brachials are the most frequently preserved
ossicles. The wing-like processes are rarely damaged suggesting that these are the most robust plates
in the crinoid skeleton. These heavily calcified ossicles are incongruous with an endoskeleton that

MILSOM: CRINOID SACCOCOMA

125

supposedly has been reduced to its minimum weight. The thickness of the radial plates which
enclose and protect the visceral mass has been reduced to a minimum; therefore it seems unlikely
that an animal would reduce the weight of such vital plates whilst strengthening those which provide
support for a supposed connecting membrane.

Presence of a soft tissue membrane

Although the curved margin of the wing would maximize the area of attachment of the membrane,
the wing surface is completely smooth and the curved margins are an unusual shape for membrane
attachment. There is no marginal groove or surface ornament to suggest connection to a soft tissue
membrane in life. Wing size decreases distally, which is also inconsistent with the support theory.
It would be more logical for the area of support required and the radius of the membrane to be
directly proportional. However, they are inversely proportional in Saccocoma (Text-fig. 3). Rather

text-fig. 3. Decrease in flange size with an increase in
area of the inferred membrane.

than supporting a connecting membrane, for which there is no direct evidence, the brachial wings
may have acted as weight distributors performing a function similar to that of a snow-shoe,
preventing the arms from sinking into the substrate.

Thinness of the radial plates

The extreme thinness of the radial plates was cited by Jaekel (1892) as evidence of weight reduction
associated with a pelagic lifestyle. However, as the radials enclose the visceral mass, thinning the
plates would make the animal extremely vulnerable to predators, particularly if lifestyle was pelagic.
If Saccocoma lay on the sea floor with the calyx embedded in the sediment, anchored by the reticular
sculpture on the radial exterior, the calyx would have been better protected.

Pattern of arm branching

The saccocomids exhibit an unusual pattern of arm branching. Dichotomy occurs on the second
brachial, the arms then remain atomous until approximately IIBr17 after which they branch
holotomously. Essentially, Saccocoma has ten simple arms which feather distally. This arm
branching pattern may be related to a unique harvesting mechanism. Cowen (1981) stated that arm
branching patterns in crinoids are usually related to harvesting style in different environments. If
Saccocoma fed passively by gravitational settling of particles the feathered arm sections could form
an enclosed collecting bowl and thus an effective feeding net.

126

PALAEONTOLOGY, VOLUME 37

PROPOSED MODE OF LIFE FOR SACCOCOMA

The mode of life proposed for Saccocoma differs fundamentally from the conventional theory in
that the general lifestyle is taken to be benthic. Evidence for a benthic mode of life is primarily based
on the specific gravity of Saccocoma and the absence of a plausible buoyancy or swimming
mechanism. Therefore it is proposed that Saccocoma lay unattached on the sea floor with the calyx
embedded in the sediment and the oral surface close to the sediment/ water interface (Text-fig. 4).

The proximal parts of the arms lay on the sediment surface and were stabilized by the lateral flanges
which distributed the weight and prevented the animal from sinking. They would also provide
leverage for elevating the distal, feeding parts of the arms. With the calyx anchored in the
substratum, the mouth orientated upwards and the flanged proximal arms arranged radially around
the calyx immobilizing the crinoid, the distal section of the arms was free to fulfil the food-gathering
functions.

A ‘collecting-bowl' feeding posture is postulated for the saccocomids. In the region IIBrg to
IIBr17 (between the first parallel-flanged brachial and the most distal secundibrachial before the
holotomously branched area), the arms arched upwards. By extending the ramuli outwards, an
effective feeding net was formed. Due to the flexibility of the arms, there was the facility for
rheophilic feeding when a slight current was present. Under such circumstances the arms could have
orientated themselves with the aboral arm surface facing down current, as observed by Meyer
(1973) in living comatulids.

The extreme flexibility of the arms, indicated by the extent to which they could coil without
rupture, suggests that the crinoid may have been mobile. Saccocoma was probably able to crawl
across the substrate, aided by extension and contraction of the arms in a similar manner to living
comatulids (Shaw and Fontaine 1990). The extended ramuli in the holotomously branched area
would produce a large surface area to 'push off’ the substrate and, as the arm lifted away from the
substrate flattening the ramules against the main arm axis, would minimize drag and thus increase
crawling efficiency.

It is also likely that saccocomids had the ability to swim. However, due to high specific gravity,
the large amount of energy required to generate lift presumably inhibited spontaneous swimming.
Therefore, swimming may have been elicited by current action which lifted the crinoid from the
substrate, or, in extreme cases, may have been used as an escape mechanism.

LIMITATIONS TO PROPOSED BENTHIC MODE OF LIFE

A limiting factor to the proposed lifestyle is the feeding mechanism. The theory proposes that
saccocomids normally formed a ‘collecting bowl’ with which they captured food by gravitational
settling of particles. Such rheophobic feeding is extremely rare in living crinoids. Meyer (1973)

MILSOM: CRINOID SACCOCOMA

127

stated that deep-sea species of comatulids and stalked crinoids may be rheophobic, although the
collecting bowl is only a temporary feeding method exploited when currents are slack.

Although the benthic theory explains all of the major morphological characteristics, the one
feature which does not conform with the proposed life position is the angle of the articular facet.
According to the benthic theory the arms lay horizontally on the substrate perpendicular to the axis
of the calyx. However, as the radial facet is situated in the upper margin of the radial, perpendicular
to the radial ridge, the articulation with the first brachial is directed vertically rather than
horizontally. Therefore it is proposed that there was a slight proximal arching of the arms which
prevented the oral surface from sinking below the sediment/water interface.

THE PRESENCE OF SACCOCOMA IN THE SOLNHOFEN LIMESTONE

The abundance of Saccocoma in the Solnhofen Limestone remains an enigma. It is possible that as
they were unattached and free living they were easily picked up by currents and washed into the
basin. The influx of water into the basin would disrupt the salinity stratification, which is considered
the reason for the absence of a typical benthos (Seilacher 1963; Barthel 1964; Goldring and
Seilacher 1971 ; Keupp 1977), and for a time it is possible that productivity extended from the basin
floor to the water surface. Thus the saccocomids may represent opportunists exploiting a normally
hostile environment.

Many of the distinguishing morphological features of saccocomids can be interpreted as
adaptations to the basin environment. It has been postulated (Hemleben 1977) that the colonization
of the Solnhofen basin by macrobenthos was inhibited by the soupy texture of the sediment surface.
The wing-like flanges which occur on the proximal brachials of saccocomids would have supported
them on the sea floor. Similar adaptations are seen in the rare examples of accepted benthic animals
which occur in the Solnhofen Limestone. The gastropod Spinigera spinosa has two long spines,
developed on each whorl perpendicular to the longitudinal axis, which have been interpreted as soft
sediment supports (Viohl 1985).

In the quiet stable environment of the plattenkalk basins, the collecting-bowl feeding posture
would have been ideal. The arms were able to arch upwards away from the sediment surface and
thus avoid contact with the substratum and contamination by detrital sediment. The abundance of
coccoliths and other protists would provide a suitable diet, similar to that recorded for the
suspension-feeding brittlestar Ophiothrix fragilis (Warner and Woodley 1975).

It is possible that saccocomids could survive in the plattenkalk basins only for short periods of
time. However, no crawling traces attributable to Saccocoma are recorded from the Solnhofen
Limestone. If the crinoids were killed by sudden influxes of sediment it is possible that only the
tracks of animals mobile at the time of burial would be preserved. It is also plausible that the surface
prohibited crawling. If the sediment was poorly consolidated it is likely that the saccocomids were
unable to crawl and rather than being free living they tended to stay in the same location.

CONCLUSIONS

Based on morphological evidence, the mode of life of Saccocoma was benthic. Functional
morphological interpretation leads to the conclusion that Saccocoma lay passively on the basin floor
with the calyx embedded in the soft sediment. The flanges on the brachials probably acted as
supports. At approximately mid-length the arms arched upwards. The distal ramules formed an
effective multidirectional feeding baffle which could intercept currents from any direction and/or
feed on gravitationally settling food particles.

The primary mode of locomotion in Saccocoma was probably crawling. However, the poorly
consolidated sediment in the plattenkalk basins inhibited movement. Saccocoma was also able to
swim but this facility was probably only utilized as an escape mechanism.

The sheer numbers of Saccocoma in the Solnhofen Limestone suggest that it was an opportunist.
It was brought into the basin during periods of oceanic exchange which caused total mixing of the

128

PALAEONTOLOGY, VOLUME 37

basin waters and destroyed the salinity stratification. Not all the specimens brought into the basin
survived but those that did flourished on the basin floor undisturbed. Populations were ultimately
killed by sudden influxes of sediment which buried the saccocomids and preserved them fully
articulated.

Acknowledgements. This work was carried out during the tenure of Natural Environment Research Council
grant GT4/88/GS/70. The following people are gratefully acknowledged: Professor C. R. C. Paul (Liverpool
University) for helpful discussions and reading the manuscript. Miss H. E. Clark (Liverpool John Moores
University) for drafting the diagrams. Dr R. P. S. Jefferies (The Natural History Museum, London) for
translating the German references, and Mr K. Veltkamp (Liverpool University) for the SEM photographs.

RELERENCES

agassiz, j. l. r. 1835. Prodrome d'une Monographic des radiaires ou Echinodermes. Societe des Sciences
Nature lies, Neuchdtel, Memoires, 1, 168-199.

arkell, w. j. 1947. The geology of the country around Weymouth , Swanage, Corfe and Lulworth. Memoirs of
the Geological Survey of Great Britain, London, 386 pp.
barthel, k. w. 1964. Zur Entstehung der Solnhofener Plattenkalke (unteres Unterithon). Mitteilungen der
Bayerisclie Staatssammlung fur Paldontologie und Historische Geologie , 4, 37-69.

— SWINBURNE, N. H. M. and conway morris, s. 1990. Solnhofen : a study of Mesozoic palaeontology.
Cambridge University Press, Cambridge, 236 pp.

bather, F. a. 1924. Saccocoma cretacea n. sp. A Senonian crinoid. Proceedings of the Geologists' Association ,

35, 111-121.

cowen, r. 1981. Crinoid arms and banana plantations: an economic harvesting analogy. Paleobiology , 7,
332-343.

denton, E. J. 1974. Buoyancy in marine animals. Oxford University Press, London, 15 pp.
goldfuss, G. a. 1826-33. Petrefacta Germaniae 1, 115-221, Arnz, Dusseldorf.

goldring. r. and seilacher. a. 1971. Limulid undertracks and their sedimentological implications. Neues
Jahrbuch fur Geologie und Paldontologie. Abhandlungen, 137, 422^142.
hemleben, c. 1977. Autochthone aus aullochthone Sedimentanteile in den Solnhofener Plattenkalken. Neues
Jahrbuch fur Geologie und Paldontologie. Monatshefte, 1977, 257-271.
hess, h. 1972. Planktonic crinoids of Late Jurassic age from leg 1 1 deep sea drilling project. Deep Sea Drilling
Project, Initial Report, 11, 631-634.

jaf.kel, o. 1892. Uber Plicatocriniden, Hyocrinus und Saccocoma. Zeitschrift der Deutschen Geologischen
Gesellschaft, 44. 619-696.

keupp. h. 1977. Ultrafazies und Genese der Solnhofener Plattenkalk (Oberer Malm, Stidlichen Frankenalb).

Abhandlungen der Naturhistorischen Gesellschaft zu Niirnberg, 37, 128.
manni, R. and Nicosia, u. 1984. Crataegocrinus toneiellii n. gen. n. sp. from the Dogger of the Central Appenines.
Bollettino della Societd Paleontologia Italiana , 23, 137-139.

- 1985. Saccocoma vernioryi n. sp., a new saccocomid from the Upper Jurassic of Central Italy.
Geologica Romana, 23. 89-97.

Meyer, d. l. 1973. Feeding behaviour and ecology of shallow-water unstalked crinoids (Echinodermata) in the
Caribbean Sea. Marine Biology. 22, 105-129.

orbigny, a. d. d'. 1852. Prodrome de la paleontologie stratigraphicjue universelle des animaux mollusques et
rayonnes faisant suite an cows elementaire de paleontologie et de geologie stratigraphique, 3. Victor Masson,
Paris, 1-189.

PECK, R. E. 1955. Cretaceous microcrinoids from England. Journal of Paleontology, 29, 1019-1029.
QUENSTEDT, f. a. 1852. Handbuch der Petrefactenkunde. H. Laupp Verlag, Tubingen, 792 pp.
seilacher, a. 1963. Umlagerung und Rolltransport von Cephalopoden-Gehausen. Neues Jahrbuch fur Geologie
und Paldontologie. Monatshefte. 1963, 593-615.

reif. w. e. and westphal, f. 1985. Sedimentological, ecological and temporal patterns of fossil
Lagerstatten. Philosophical Transactions of the Royal Society of London , Series B, 311, 5-23.
shaw, G. D. and fontaine, a. r. 1990. The locomotion of the comatulid Florometra serratissima
(Echinodermata: Crinoidea) and its adaptive significance. Canadian Journal of Zoology . 68. 942-950.
sieverts-doreck, h. 1955. Die Verbreitung der Crinoidengattung Saccocoma im Schwabischen Jura.
Jahreshefte des Vereins fuer V aterlaendische Naturkunde in Wuerttemberg , 110. 118-120.

MILSOM: CRINOID SACCOCOMA

129

verniory, r. 1960. Presence (et varietes) de Saccocoma tenella Goldfuss a Talloires (Haute-Savoie). Archives
des Sciences, Geneve, 14, 250-257.

— 1961. Presence de Saccocoma cjuenstedi Dreck (in coll.) dans les gorges de la Meouge (Sisterton-Provence).
Archives des Sciences, Geneve, 14, 315-320.

1962. Quelques considerations sur les Saccocomides (echantillonnage, statistique, stratigraphie). Une
nouvelle forme de Saccocoma (Montbrand Hautes-Alpes, France). Archives des Sciences, Geneve, 15,
388-391.

viohl, G. 1985. Geology of the Solnhofen lithographic limestone and the habitat of Archaeopterx. 31-44. In
hecht, m. K., ostrom, j. h., viohl, g. and wellnhofer, p. (eds). The beginnings of birds. Proceedings of the
international Archaeopteryx conference, 1984. Freunde des Jura-Museums, Eichstatt, 394 pp.
walther, j. 1904. Die Fauna der Solnhofener Plattenkalke bionomische betrachtet. Festschrift der Medizinisch-
naturwissenschaftlichen Gesellschaft zu Gena, 11, 133-214.

Warner, G. f. and woodley, j. d. 1975. Suspension feeding in the brittle star Ophiothrix fragilis. Journal of
the Marine Biological Association of the United Kingdom, 55, 199-210.
zittel, k. a. von. 1876-80. Handbuch der Paldontologie , Band I Palaozoologie. Miinchen and Leipzig,
vii + 765 pp.

CLARE V. MILSOM

School of Biological and Earth Sciences
Liverpool John Moores University

Typescript received 4 July 1992 Byrom Street

Revised typescript received 29 March 1993 Liverpool L3 3AF, UK

GROWTH AND DISINTEGRATION OF BIVALVE-
DOMINATED PATCH REEFS IN THE UPPER
JURASSIC OF SOUTHERN ENGLAND

by FRANZ T. FURSICH, TIMOTHY J. PALMER
Cind KAY L. GOODYEAR

Abstract. Patch reefs, up to 4 metres high and 8 metres across, grew amongst oolith shoals at the top of the
Portland Limestone Formation (Portlandian, Upper Jurassic) on the Isle of Portland, southern England.
Principal reef framebuilders, which provided between 55 and 70 per cent of the reef volume, were cementing
bivalves, solenoporacean algae, and bryozoans. The remaining pore-space in the reef was filled by sediment,
most of which is in the form of a precipitated peloidal cement. The cement lithified the reef while it was still
exposed on the sea floor, and was probably precipitated under bacterial control. A diverse accessory fauna of
small cementing encrusters and nestlers includes groups such as terebratulid brachiopods and lithistid sponges
that have not previously been found in the Portland Limestone. Serpula ( Cycloserpula ) striatissima sp. nov. and
Carterochaena pulcherrima gen. et sp. nov. are described. Both the primary organic framework of the reef and
the submarine cements were bored by a variety of endohths, which locally removed as much as 40 per cent of
the reef volume. Vacated borings acted as sites for precipitation of further peloidal cement. Borings are well
preserved as natural three-dimensional casts in cases where they originally perforated an aragonite substrate
which has since dissolved. New taxa of borings consist of Cunctichnus probans ichnogen. et ichnosp. nov.,
Spirichnus spiralis ichnogen. et ichnosp. nov., Talpina bromleyi ichnosp. nov., and Entobia cervicornis ichnosp.
nov.

The Portlandian limestones of Dorset in southern England constitute some of the most closely-
scrutinized carbonate sections in the world, given their place in the field-trip itineraries of many
educational establishments, commercial operations, and local interest groups. They are excellently
developed and exposed on the Isle of Portland in Dorset, southern England (Text-fig. 1 ), where the
topmost Portlandian is developed in an oolitic freestone facies that has traditionally been regarded
as England’s best building stone for fine quality ashlar and mouldings. Hence there are many good
quarry exposures on the island as well as natural outcrops in the cliffs. Notwithstanding their fame,
however, the Portland limestones in Dorset have received only a modest amount of scientific
attention. Nineteenth and early twentieth century studies described stratigraphical, lithological and
palaeontological details (Buckland and De la Beche 1836; Fitton 1836; Blake 1880; Damon 1884;
Woodward 1895; Arkell 1947) and the current state of biostratigraphical refinement within the
Portlandian Stage has only recently been attained (Wimbledon and Cope 1978; Wimbledon in Cope
et al. 1980). The only modern sedimentological study that deals with depositional processes and
environments is that of Townson (1975). Surprisingly, it was not until this work that the spectacular
patch reefs that form the subject of this paper were first recognized as a distinct facies. This is
probably a reflection of the relative inaccessibility of the Isle of Portland in former times, the very
local development of the reef facies in the more northerly part of the island where most of the
quarries were concentrated last century, and its former lack of economic value. This study thus sets
out to describe the palaeoecology and the sedimentary dynamics of these spectacular but little-
known build-ups, whose guild structure shows many similarities with modern, coral-dominated
patch-reefs.

(Palaeontology, Vol. 37, Part 1, 1994, pp. 131-171, 7 pls.|

© The Palaeontological Association

132

PALAEONTOLOGY, VOLUME 37

text-fig. 1 a, sites on the Isle of Portland, Dorset,
where patch-reefs have been examined for this study:
1, Inmosthay Quarry; 2, Coombefield Quarry;
3, Suckthumb Quarry; 4, Perryfield Quarry; 5, shore
north-east of Portland Bill, b, location of Isle of
Portland on the coast of southern England.

STRATIGRAPHICAL AND GEOGRAPHICAL FRAMEWORK

Lithostratigraphy and age

The reefs occur in the top 3 to 4 metres of the Portland Stone, in the upper part of the Portland
Freestone Member of Cope et at. (1980). Townson (1975) called this unit the Winspit Member of
the Portland Limestone Formation, but use of this name has not caught on.

All the individual beds in the upper part of the Portland Stone on Portland (from where nearly
all the building stone comes) have individual quarrymen's names which are widely used by
geologists, and which have appeared in many earlier geological accounts (e.g. Arkell 1933, 1947).
The topmost oolitic grainstone widely becomes fossiliferous towards its top, giving rise to a horizon
(not usually more than about 0-7 m thick) called the Roach, which is particularly characterized by
preservation of open biomoulds after aragonite. The familiar Roach contains a soft-sediment fauna
dominated by high-spired gastropods and infaunal bivalves. At a few localities, the Roach thickens
to as much as 4 m at the expense of the rest of the bed and develops the biohermal character that
we describe here. The Portland Freestone Member belongs to the Zone of Titanites anguiformis
(Cope et at. 1980).

The Roach Bed in the Portland Quarries is immediately overlain by the palaeosols and
thrombolites of the basal Purbeck Limestone Formation (Francis 1986), indicating emergence from
the marine, into subaerial and lacustrine conditions.

Patch-reef localities

Renewed interest in natural building materials in the last decade has seen the reopening and
expansion of many of the older quarries on Portland, and both small and large examples of the
patch-reefs have been exposed and examined by us as quarry walls have receded. The reef facies is

FURSICH ET AL.: JURASSIC REEFS

133

sometimes slabbed and used commercially as a cladding stone on modern buildings, so that large
blocks of the native reef rock may sit around in the quarries awaiting sale after the natural outcrop
has been quarried away. In situ reefs in existing quarry walls can be seen at Inmosthay and
Suckthumb Quarries and in the coastal exposure about 1 km northeast of Portland Bill. A very large
example was formerly exposed at one of the quarries east of the main road in Easton (D. W. J.
Bosence pers. comm.). We have also studied the reef facies in Perryfield and Coombefield quarries.
During 1990-1992, Coombefield was expanding very rapidly and at least 20 reefs were examined.
These localities are marked on Text-figure 1. As mentioned above, the reef facies seems to be more
widely developed in the southern part of the island. Reefs are unknown in the Portlandian rocks
of the mainland.

STUDY METHODS

We have studied in situ reefs and quarried blocks derived from them. Data on shape, composition, orientation,
and lateral passage into flanking sediments have been obtained in the field, and an extensive fauna has been
collected. Large, orientated slabs have been cut and polished on a vibrating lap, yielding a wealth of
information about compositions and textures that cannot be seen on unpolished material. Data on proportions
of different reef components and biovolumes of the main framebuilders have been obtained from these surfaces
by overlaying them with a 5 mm grid marked on a transparent overlay and point counting at the intersections.
Amounts of material removed from the reef by small borers were obtained by point counting on acetate peels
of polished surfaces viewed through a petrological microscope. Bulk samples of the reef were broken up, sieved,
and picked until several examples of even the rarest elements of the fauna had been collected or observed.

Repository. Collections are housed in the Oxford University Museum.

ENVIRONMENTAL FRAMEWORK
Colonization of the Upper Portlandian sea-floor

The abundant benthic fauna in the Roach Bed suggests that the oolith shoals of the Portland
Freestone Member did not remain actively growing right up until the onset of Purbeck deposition,
but passed into a stable phase when colonization occurred. The heavily-micritized state of the
ooliths supports the idea of a prolonged period of stable bottom conditions after the main episode
of shoal growth, and the diversity of both infauna and epifauna would not have grown on
constantly shifting, unstable, active oolith shoals.

The normal fauna of the Roach is dominated by soft-sediment dwelling bivalves (particularly
Laevitrigonia) and gastropods ( AptyxieUa ), which may occur together or in assemblages dominated
by one or the other. Locally, small (c. 10 mm across) Solenopora , encrusting small shell fragments,
are prevalent. A fourth variation contains conspicuous Camptonectes and Isognomon (Text-fig. 2).
Many of the bivalves are found articulated and there is little sign of current sorting or abrasion,
though locally there are signs of current orientation of single valves. However, both gastropod and
bivalve shells, including those of the burrowing species, are encrusted by an epifauna of bryozoans
and oysters, which often attach to the inner faces of the valves. This suggests that dead shells lay
on the sea floor for some time before eventual burial. Locally, these shelly layers supported
Camptonectes and. particularly. Isognomon in great profusion. Again, they are usually articulated,
yet are encrusted inside and out. In life, they must have formed extensive banks. These observations
indicate that conditions on the sea-floor were not subject to any great current activity (though
probably intermittent storms caused local shell orientation), and were conducive to growth of
abundant life. We agree with Townson’s ( 1975) view of a fully marine environment in shallow water
on top of a prograding carbonate platform, in which organic growth was the principal sedimentary
process. Thus, once the formerly active oolith shoals had stopped shifting around, skeletal growth
became the dominant sedimentary process and their history was one of colonization by a benthos,
with gradual accumulation of shells and occasional episodes of winnowing leading to a change in

PALAEONTOLOGY, VOLUME 37

active oolite shoal

I I

STABILISATION

g inactive stable substrate

I — I

COLONIZATION

(BIOPRODUCTION » SEDIMENTATION RATE)

seeding of daughter
reefs by erosion of
large reef blocks

bivalve-dominated reefs

Isognomon banks

EROSION AND/OR SEDIMENTATION

U

•— J— • TERMINATION OF REEF GROWTH >-j-«

-fig. 2. Generation of shell pavements on which Portland patch-reefs initiated and grew, as a result of
episodic winnowing of the fauna of the inactive shoals of oolitic lime sand.

FURSICH ET AL.: JURASSIC REEFS

135

the character of the sea-floor. This taphonomic feedback (Kidwell and Jablonski 1983) resulted in
gradual replacement of a soft-sediment infaunal benthos by a shell-pavement epibenthos. Events are
summarized in Text-figure 2.

Reef growth and composition

It was against this sedimentary background that, locally, successive generations of encrustation by
a variety of frame-building organisms started to build up biocemented frameworks which, in turn,
attracted further encrustation and developed into small reefs. The smallest such buildups are only
a few tens of centimetres across, but locally the process went on for a long time, eventually resulting
in bioherms as large as 7 m across and 3-5 m high. A section through the largest reef body exposed
during the course of this study is shown in Text-figure 3. It is difficult to get a clear picture of the

n

REEF

FACIES

]REEF DEBRIS 1° o I OOLITIC SAND [— ] I POCKETS OF

J FACIES 1—2 — 3 FACIES L—J LIME MUD

CHERT

0 Metre 1

text-fig. 3. Facies distribution in a vertical section taken through a large patch-reef in the north-east corner

of Suckthumb Quarry (see Text-fig. 1).

three-dimensional shapes of many of the reefs because of the two-dimensional exposure and the
rubble surrounding them in the working quarry faces. The strength of the reef fabric initially came
from the principal framebuilders, but the porosity within the frame acted as a site for the
precipitation of large volumes of peloidal calcite marine cement, which rapidly converted the
primary organic structure into a dense reef-rock (PI. 1, figs 1-2; PI. 2, figs 1-2). This peloidal fabric
is often laminated, due to variations in peloid size from layer to layer (PI. 2, figs 1-2), and fills some
of the primary cavities geopetally. It is structurally identical to early marine cements that have been
described from a variety of reefal settings from ancient (see references in Sun and Wright 1989, and
Mock and Palmer 1991) and Recent (Macintyre 1985) settings, and for which some workers have
proposed a bacterially-mediated origin (Chafetz 1986).

The peloidal cement fills smaller spaces between the main framebuilders, and is particularly
abundant in borings that perforate the primary fabric, and inside articulated shells (PI. 2, fig. 2).
Larger intrareefal cavities locally contain detrital sediment, such as oolilhs and unpelleted lime mud
which is burrowed (Text-fig. 3). Relative proportions of the different components differ within the
reef body, and around the edges of reef blocks where active reef growth had ceased and where
bioerosional processes dominated (PI. 2, fig. 3). In this distinct edge sub-facies, many generations
of boring, precipitation of peloidal cement in the borings, and reboring alternated. As a result the
edge facies has a significantly higher proportion of marine pelleted cement than is found deeper into
the reef. These differences in the percentage volume contributions of the different organic and

136

PALAEONTOLOGY, VOLUME 37

inorganic constituents of the reefs, which were obtained by point counting as discussed above, are
summarized in Text-figure 4.

Relationship of reefs to adjacent soft sediment

Oolitic sediments abut against the sides of the larger reef bodies and locally fill voids within the reef
structure (Text-fig. 3). Smaller reefs are covered by such sediment, which must have accumulated
at a rate greater than reef growth, or else have washed over them during occasional high energy
events. The largest reefs intersect the top of the Portland and are directly overlain by Purbeck
sediments.

There is plenty of evidence for erosive processes carrying material off the reefs into the adjacent
oolitic sands. Immediately next to the reefs, many large shells of reef-dwelling and frame-
constructing species are found in the oolitic matrix (Text-fig. 5). Further away from the reef,
abundances of reefal groups and the sizes of the derived fragments decrease, so that more than
about 20 m away there is no indication of the proximity of the reef, and the normal soft-sediment
Roach assemblages are seen. We have not been able to see whether this detrital apron extends
farther in one lateral direction than any other, which might suggest a windward and a seaward
direction. Small blocks of the primary frame, apparently broken off' larger reefs and lying close to
them, provided sites for further encrustation and growth (Text-fig. 3). Sometimes these daughter
reefs subsequently coalesced with the parents that spawned them.

LATER DIAGENESIS

All facies of the Portland Roach are characterized by complete leaching of aragonite, so that all
aragonitic taxa are preserved as internal and external moulds (PI. 2, fig. 1). The dissolution probably
occurred very soon after completion of sedimentation when the sequence was uplifted into the
meteoric realm. The overlying basal Purbeck rocks contain soils and freshwater ostracodes (Arkell
1947 ; Townson 1975; R. C. Whatley pers. comm.) so downward flow of meteoric water through at
least the upper part of the Portlandian sediments must have occurred, and would probably have
been rapid at some times during the year, given the marked seasonality of Lower Purbeck times
(Francis 1984). In contrast to the fate of aragonite, low-magnesian calcite shells of bryozoans,
brachiopods, some molluscs, and the winter layers (cf. Wright 1985) of the rhodophyte Solenopora
are preserved with full microstructural detail, and show only some mild silicification (presumably
the result of remobilization of silica from sponge spicules; Townson 1975). However, a third style
of preservation is evident in serpulids, encrusting forams ( Nubeculinella ), and in the darker summer
layers of Solenopora. In these taxa preservation is variable, ranging from good (with little or no
apparent structural alteration) to mouldic (with complete loss of original shell; PI. 1, fig. 2). In
between (particularly in Solenopora ) lie examples of partial replacement by diagenetic calcite with
varying amounts of loss of detail of the original microfabric. This style of replacement has been
noted in Solenopora of Middle Jurassic age by Wright (1985) who favoured the explanation that
composition fluctuated between low- and high-magnesian calcite (LMC and HMC) across the
seasons and that the latter was replaced by aragonite early in diagenesis. Given that the degree of
dissolution is never as clear-cut as in the unequivocally aragonitic shells, we support an original

EXPLANATION OF PLATE 1

Figs 1-2. Two pieces of Portland patch reef showing three of the four principal framebuilders: Solenopora (S),
Plicatula (P), and Liostrea (L), surrounded by sedimentary matrix which is either precipitated micrite
peloidal cement (C) or oolitic grainstone that has been washed into the reef from the surrounding oolite
shoals (O). Large borings (B) containing articulated Lithophaga cut across both the skeletal frame and the
synsedimentary cement. Polished blocks. Large squares on scale grids are 10 mm.

PLATE 1

FURSICH et a /., Portland patch reef

138

PALAEONTOLOGY, VOLUME 37

HMC composition for Solenopora and the other variably-preserved skeletons, with magnesium
levels possibly varying throughout the year in response to temperature or growth rate. Styles of
dissolution of HMC depend on a range of factors such as microstructural detail, the original
magnesium levels in the calcite, the concentrations of dissolved carbonate in the diagenetic fluids,
and the evolution of the pore-water chemistry with increasing burial (Walter 1985). Variations in
these probably led to the differences in the degree of alteration of putatively HMC skeletons that
are now preserved, and we see no reason to support an additional diagenetic episode involving
differential replacement of HMC by aragonite.

It is not so much the loss of aragonite that makes the Roach remarkable, but the fact that the
moulds were not filled by later diagenetic calcite cement. There is a small amount of this cement,
mostly growing as a thin sugary veneer on some of the aragonite moulds and in the small pore
spaces inside borings or between the peloids of the early marine cements, but it is volumetrically a
minor component. It may be that, the aragonite and some of the HMC having been flushed out of
the system during meteoric emergence at the end of Portland times, there was no major source of
dissolved carbonate when the sequence eventually became buried. Indeed, the whole Portland
sequence may have been diagenetically closed, sandwiched between the underlying Kimmeridge
Clay and the overlying clay-rich Purbeck and Wealden.

REEF ECOLOGY: GROWTH, COLONIZATION AND BIOEROSION
Ecological guilds

We have recognized five ecological guilds into which all of the fauna of the Portland reefs can be
placed. These provide a convenient structure for a discussion of the dynamics of reef growth
and destruction. These guilds are: (1) the principal organic framebuilders; (2) the accessory
framebuilders of small cementing encrusters; (3) the interstitial fauna of nestlers and predominantly
soft-boiled sessile forms; (4) vagile strollers on the reef surface; and (5) the borers that excavated
dwelling holes in the skeletons of the primary framebuilders and in the cemented reef rock. A
summary of the complete reef and reef-associated fauna, together with the guild assignations, is
given in Table 1.

Principal framebuilders

There are four major taxa that are responsible for the biogenic frame of the reefs. Most conspicuous
is the red alga Solenopora ‘ portlandica ’ (PI. 1, figs 1-2), which ranges in size from small knobs only
about 10 mm across to large cauliflower-shaped heads as much as 0-3 m high. (This species has yet
to be formerly described, although it is widely known in the literature (e.g. Wright 1985) under this
name.) Growth banding is very conspicuous and the largest specimens display as many as fifty
alternations of lighter (winter) and darker (summer) layers. However, although Solenopora is locally
very common and conspicuous in some of the reefs, elsewhere it may be rare or absent altogether.

The framebuilders that are most widespread, and which make the biggest contribution in terms
of volume to all the reefs examined, are bivalves. The first is Liostrea expansai. Sowerby (Text-fig. 5),
a thick-shelled oyster that grows in a somewhat cup-shaped form (probably as a result of
crowding). The range of shape and size variation is difficult to assess because the shells cannot easily

EXPLANATION OF PLATE 2

Figs 1-2. Structure of reef rock from the Portland reefs. 1, polished slab showing primary biogenic fabric of
oysters (dark laminated shells) and Plicatula (with biomouldic preservation) surrounded by peloidal micrite
cement. Both are cut by bivalve borings, which are in turn filled with later generations of laminated peloidal
cement ; x 08. 2, acetate peel of laminated geopetal peloidal early marine cement inside bivalve boring; x 18.

Fig. 3. Large Lithophaga boring into edge of patch reef ; x 06.

PLATE 2

FURSICH et al Portland patch reef

140

PALAEONTOLOGY, VOLUME 37

7.3%

REEF BODY

35.7%

7,021 points counted
over 3,51 1 sq.cm

Principal Framebuilders

Minor Framebuilders

Pelleted Micrite
between framebuilders
and in borings

Others

REEF EDGE

58.0%

1 ,262 points counted
over 631 sq.cm

text-fig. 4. Different proportions of components in the main reef body and at the edges of mature reefs. The
greater amount of peloidal cement in the latter is mostly located within borings, which occupy a larger

proportion of the overall volume at the reef edge.

be extracted whole from the reef. The second bivalve is Plicatula damoni Cox (PI. 2, fig. 1) which
is easy to study because the valves were composed of a thick (c. 5-6 mm) inner aragonite layer and
a thin exterior layer of calcite folia. The former has dissolved during diagenesis, freeing the internal
moulds which are largely composed of peloidal cement. Because of crowding and overgrowth, they
are very variable in shape and size, and attain a maximum height of 50 mm.

The fourth principal framebuilder, which is only present in some of the reefs, is the massive
multilamellar cyclostome bryozoan Hyporosopora portlandica (Gregory). Uniquely among Jurassic

FURSICH ET A L.\ JURASSIC REEFS

141

. , - '
^ • -■ - > C

_ /

text-fig. 5. Junction between reef
and overlying oolitic sediment con-
taining skeletal debris that was
washed off the reef. In this sample, the
reef is largely composed of massive
bryozoans (Hyporosopora portlandi-
ca) with embedded oysters. Scale-bar
represents 10 mm.

Bryozoa, what are apparently single colonies can reach as much as a metre across and several tens
of centimetres thick. Polished vertical sections show lighter and darker zones of about five zooecial
layers in thickness, which reflect differences in zooecial size that may be seasonal. Bivalved oysters
are widely embedded in these massive bryozoan colonies (Text-fig. 5).

We have studied the relative contributions of the four main framebuilders to the buildups in two
ways. In the field, we counted numbers of individuals of the four types on known areas of vertical
sections through four different reefs at two localities (Text-fig. 6). The results show some of the
variation, particularly in the abundance of Solenopora and Hyporosopora , but numbers alone do
not give a clear indication of the contribution of each species to reef growth. We thus studied
biovolume by point-counting cut and polished surfaces of twenty three blocks collected from several
different reefs in Coombefield Quarry, subjectively distinguishing between the reef body and the reef
edge sub-facies. The total area examined in this way was over 0 64 nr. The blocks were chosen for
appropriate size rather than by composition, so the proportions of the different framebuilders are
likely to reflect actual proportions in the reefs thus sampled. However, some of the blocks were lying
loose in the quarry, so it is not possible to say how many reefs are represented. Results of this
exercise are shown in Text-figure 7, and clearly show that reef growth is dominated by the bivalves.

142

PALAEONTOLOGY, VOLUME 37

table 1. Summary of species in the five reef guilds of the uppermost Portlandian, Isle of Portland, Dorset. Key:
+ , trace fossil; + +, bioimmuration; x, new record or first assignation of this taxon to the British
Portlandian; ( x ), species previously recorded, but new material gives more detailed morphological
information.

Principal framebuilders

Algae

Solenopora ' portlandica ’

Bivalvia

Plicatula damoni Cox
Liostrea expansa (J. Sowerby)

Bryozoa

Hyporosopora portlandica (Gregory)

Accessory framebuilders

Stomatopora sp.

Nubeculinella sp. x

Porifera

lithisthid indet. x

‘Worms’

Serpula (Cy closer pula) gordialis
(v. Schlotheim)

Serpula ( Cy closer pula) striatissima x
sp. nov.

Serpula ( Dorsoserpula ) sp.

+ agglutinating worm tube x

Bryozoa

' Be renicea ’ sp.

Stomatopora sp.

Interstitial fauna
Porifera
Rhaxella sp.

Pachastrella sp.

Bivalvia

Barbatia (Barbatia) bourgueti de x

Loriol

Cucullaea ( Idonearca ) sp.

Modiolus (Modiolus) sp.

Isognomon bouchardi (Oppel)

Plagiostoma sp.

Hiatella ( Pseudosaxicava ) phaseolus ( x )
(Eudes-Deslongchamps)

Bryozoa

+ + Stolonicella sp. x

Brachiopoda

micromorphic terebratulid indet. x

Incertae sedis

+ + bioimmured ‘dimples’ x

Strollers

Bivalvia

‘Lucina ’ portlandica J. de C. Sowerby
Gastropoda

Uchauxia quadrigranosa Cox
naticid indet.

medium-spired gastropod indet.
pleurotomariid indet.

Annelida

+ Arachnostega gastrochaenae x

Bertling

Borers

Thallophytes

+ unidentified filaments x

Porifera

+ Entobia cervicornis isp. nov. x

‘Worms’

+ Caulostrepsis cretacea Voigt x

+ Spirichnus spiralis igen. et isp. x

nov.

+ Cunctichnus probans igen. et isp. x
nov.

Bivalvia

Lithophaga ( Lithophaga ) x

subcyclindrica (Buvignier)

Lithophaga sp. A x

[+ Gastrochaenolites torpedo Kelly x
and Bromley]

Carterochaena pulcherrima gen. et x
sp. nov.

Gastrochaenopsis recondita (Phillips) x

[+ Gastrochaenolites dijugus Kelly x

and Bromley]

Phoronida

+ Talpina bromleyi isp. nov. x

Bryozoa

+ Iramena isp. x

Crustacea

+ Rogerella pattei (Saint-Seine) x

The dominance of Plicatula in the reef-edge sub-facies reflects the density of borings there. Plicatula
can grow in highly twisted and distorted shapes and sometimes occupies vacated bivalve borings,
or perches on pinnacles between borings. This flexibility of growth form would have made it an
effective competitor for the limited space in the growing reef. The other framebuilders seem less able
to cope with the vicissitudes of this highly competitive and bioerosive environment. Moreover, the

FURSICH ET AL.\ JURASSIC REEFS 143

FRAMEBUILDERS

Plica tula
Liostrea
Solenopora
Hyporosopora

n = 219 n = 219 n = 95 n = 474

text-fig. 6. Proportions of the four principal framebuilders (by number) in two reefs from Coombefield and

two from Inmosthay Quarries.

FRAMEBUILDERS - biovolume

text-fig. 7. Relative proportions (by volume) of the four principal framebuilders in the main reef body and
at the reef edge. Data obtained by point counting on polished blocks; see text and Text-figure 6 for details.

bivalves are likely to have been faster growing than the more massive framebuilders, less controlled
by incident light than Solenopora , and able to produce large numbers of individuals, which often
settled and grew in bundles, in frequent recruitment episodes to take advantage of new attachment
areas as soon as they became available.

Coombefield I Coombefield

(1850 cm2) (3800 cm2)

20 60 '/. 20 60

Inmosthay 1

(2000 cm2)

20 60

Inmosthay 2

(7100 cm2)

20 60 %

I

1

rare

abundant

144

PALAEONTOLOGY, VOLUME 37

ORIENTATION

text-fig. 8. Orientations of in situ Liostrea and
Plicatula , taken from field observations of reefs in
Coombefield and Inmosthay Quarries.

Liostrea and Plicatula also exhibit different orientation patterns and hence growth strategies
(Text-fig. 8). Individuals of L. expansa grew preferentially upward, whereas P. damoni seems to have
been able to grow in any orientation. In terms of numbers of individuals, Plicatula is usually the
most abundant framebuilder and was able to occupy the sides of the reefs and overhangs from
which it grew outwards and downwards, as well as growing upwards from the reeftops. It was thus
a major contributor to reef lateral expansion as well as upward growth. In contrast, the preference
for the oysters to grow with negative geotaxis suggests that their main contribution to reef growth
was in the upward direction.

Accessory framebuilder s

The accessory framebuilders are dominated by small lamellate bryozoans, serpulids, and encrusting
forams. Runner-type bryozoans ( Stomatopora ), small oysters (including both Nanogyra and other

EXPLANATION OF PLATE 3

Figs 1-2, 4. Araclmostega gastrochaenae Bertling. 1-2, OUM J53650; latex cast of negative reliefs on the
surface of an internal mould of Plicatula ; 1, x 3; 2, x 15. 4, OUM J53652; calcite-filled endorelief at the
surface of internal mould of Plicatula damoni Cox; x2.

Fig. 3. Serpula (Cycloserpula) striatissima sp. nov. OUM J53653, holotype; latex cast of external mould; x 10.
Fig. 5. Dimpled structures, latex cast, preserved by bioimmuration on internal mould of Plicatula , OUM
J53654; x 12.

Fig. 6. Bundled scratch marks, preserved as positive relief near the adductor muscle scar of Plicatula , OUM
J53655 ; x 12.

PLATE 3

FURSICH el al Portland reef biota

146

PALAEONTOLOGY, VOLUME 37

forms which may be small L. expansa ; it is difficult to distinguish between them in cross-section)
and agglutinated tubes occur more rarely. Also occurring rarely, but as quite large individuals, is
an unidentified lithistid sponge, the first recorded from the Portlandian of England. The accessory
framebuilders as a whole are difficult to study as they are firmly embedded in the reef matrix and
cannot be extracted. The larger ones, however, can be recognized in cross-section and have been
included as a separate category in the point-counts (Text-figs 4 and 7).

There is one microenvironment in the reef where relative abundances of the different small
encrusters can easily be studied. This is inside articulated Plicatula. Here, the encrusters grew inside
at least 60 per cent of the dead shells and were subsequently smothered by sediment or marine
pelleted cement. Subsequent dissolution of the aragonite shell leaves this sedimentary filling as a
well-preserved steinkern (PI. 3, figs 1 and 4) that shows the attachment faces of the encrusters that
are embedded in it and which were originally attached to the inner faces of the valves. Alternatively,
encrusters that have themselves dissolved away leave external mouldic impressions in the steinkern
surface (PI. 3, fig. 3). Text-figure 9 shows relative numbers of Plicatula in three different reefs that
have encrusters of different species in them. Although this small cryptic microenvironment might
be expected to have a somewhat specialized encrusting fauna, we anticipate that much of the surface
space in the reefs that was available to accessory framebuilders would have been of a cryptic kind,
in limited spaces down between the main framebuilders and in the reef interior. Typical Jurassic
upward-facing, open-surface encrusters such as Nanogvra and large Dorsoserpula occur rather
rarely in the reef body.

Interstitial fauna

The most conspicuous members of the interstitial fauna are the byssate bivalves that nestled inside
bivalved shells, old borings, and other small cavities. These include the familiar arcids and mytilids
widely seen in Jurassic rocks, but there are also large numbers of Hiatella which is generally poorly
known from the Jurassic, probably because the small, thin-shelled individuals are likely to be
overlooked in facies that do not show mouldic preservation. The great majority of the Hiatella in
the Portland reefs are found nestling inside the vacated crypts of boring bivalves, sometimes inside
the gaping shells of the borers that made the holes in the first place. Indeed more than one pair of
Hiatella valves can be found, later occupants inside earlier ones. They are very variable in shape and
size (though never more than 10 mm long) and clearly modified their shape to fit the hole they
nestled in. There is no indication that they enlarged or modified this hole in any way; they merely
squatted, as did the Lower Cretaceous Hiatella described by Kelly (1980).

We also regard the worms that lived inside the sedimentary fills of dead shells and vacated borings
as part of the interstitial fauna. The burrows that they made are most clearly seen as grooves in the
surfaces of Plicatula and Lithophaga steinkerns, where their burrowing activity was deflected along
the inside face of the shell (PI. 3, figs 1-2, 4). These net-like burrow systems are extremely
widespread in fossil internal moulds generally (particularly in Mesozoic rocks, but we have seen
examples from the Ordovician), and have recently been described under the name Arachnostega by
Bertling (1992). They are discussed more fully in the systematic section below.

A major contribution to the living reef appears also to have been made by demosponges that are
now only represented by calcite casts of their originally opaline silica spicules. Thin sections through
interstitial sediment and patches of pelleted cement show an extraordinary abundance of the tiny
bean-shaped microscleres (selenasters) of Rhaxella , together with lesser numbers of the tetraxon
megascleres of Pachastrella. Rhaxella is closely related to the Recent Placospongia (Hinde
1887-1912) which is common in Recent coral reefs in the western Atlantic (Riitzler 1978; Riitzler
and Macintyre 1978).

There are two additional minor curiosities among the interstitial fauna. The first is the first
record from the Portland Limestone of terebratulids, albeit very rare and very small. The second
is the preservation by bioimmuration (see Taylor 1990) of minute dimpled structures, consisting of
hemispherical mounds c. 0-25 mm across, with a little hole in the top (PI. 3, fig. 5). In some, this hole

FURSICH ET AL. \ JURASSIC REEFS

147

Inmosthay

n = 20

text-fig. 9. Patterns of en-
crustation of inner faces of
articulated Plicatula by the most
common of the accessory frame-
builders in two reefs from
Coombefield and one from In-
mosthay.

appears to have been enlarged, as if by emergence of an inhabitant. We suspect that they are eggs
of an unknown invertebrate.

Vagile fauna

‘Strollers’ are limited to gastropods, of which the tiny procerithiid Uchauxia quadrigranosa Cox,
which we envisage as a grazer and scavenger with catholic dietary tastes, was by far the most
common. The absence of echinoids, which are such a prominent member of this guild in Recent reefs
and in other Jurassic reefal settings (e.g. Fiirsich 1977; Palmer and Fiirsich 1981), remains
something of a mystery.

Borers

The feature of the Portland reefs that first drew our attention to them was the variety and
abundance of the borings. They occur both in the overall reef-rock framework and in individual
shells. They are easiest to study where they entered Plicatula valves, and later became filled by

148

PALAEONTOLOGY, VOLUME 37

precipitation of the early pelleted cement. Exquisite natural cement casts have been revealed by the
subsequent aragonite dissolution.

The largest and most conspicuous borings are examples of Gastrochaenolites torpedo Kelly and
Bromley that may reach 60 to 70 mm in length. They were made by Lithophaga subcylindrica
Buvignier, and are most common in the cemented reef-rock, particularly in the reef edge facies
where active expansion of the reef had stopped and bioerosive processes became predominant
(PI. 2. fig. 3). Locally, they cross empty cavities such as the spaces inside articulated shells, so it
seems the boring bivalve was not put of!' by breaking through into a void, but continued through
to the wall on the far side. This behaviour was not shown by either of the other boring bivalves,
which are both gastrochaenids.

Even more numerous than Lithophaga crypts are borings of the bivalves Carterochaena and
Gastrochaenopsis which differ from those of Lithophaga by being smaller, and having a circular
cross-section in the lower half but a figure-of-eight cross-section in the upper part. This is the result
of a secreted lining to the mouth of the boring, isolating the shell inside, except for the holes for the
siphons to pass through. Many Recent gastrochaenids show such a structure, made of aragonite and
secreted by the siphonal walls (Carter 1978). However, the lining is preserved in our material (in the
borings of both species), and does not show the mouldic preservation of aragonite. We think that
it was either secreted calcite, or made of a paste of very fine calcite particles released in boring.

The borings themselves can be accommodated in the ichnospecies Gastrochaenolites dijugus Kelly
and Bromley. Those of Carterochaena reach a length of about 12 mm and appear limited to the
primary framebuilders, particularly the valves of the bivalves. Gastrochaenopsis reaches nearly
20 mm and is particularly common in the cement-rich reef edge sub-facies, alongside Lithophaga.

Apart from the fairly large bivalve borings, there occur a number of smaller, more delicate
borings which have been created by members of several phyla ranging from sponges ( Entobia ),
phoronids ( Talpina ), crustaceans ( RogereUa ) to bryozoans (Iramena) and various kinds of ‘worms’
( Caidostrepsis , Cunctichnus, and Spirichnus). Despite their small size (most of these borings have a
diameter of 1 mm of less), they are very conspicuous, because they either form extensive three-
dimensional ramifying systems ( Entobia ) or else occur in high density (e.g. bundles of Talpina and
Spirichnus). They are very conspicuous and occur profusely with exquisite preservation as natural
casts in the voids after dissolution of the aragonite of the Plicatula valves (PI. 4, fig. 1). They also
occur with similar densities in the oysters, and the rigidity and mechanical strength of many of the
bivalve framebuilders must have been severely weakened.

Counts of borings in more than two hundred Plicatula shells from four different reefs showed that
between 60 and 100 per cent of the shells contained at least some borings (Text-fig. 10).
Gastrochaenolites and Entobia were most abundant, followed by Spirichnus , Talpina , Cunctichnus ,
and Caulostrepsis. Of least importance were the acrothoracican barnacle boring RogereUa and
ctenostome bryozoan borings. In order to get an estimate of the relative importance of the various
borers as destroyers of the reef framework, which cannot be judged from their relative numbers
alone, point counts were made on acetate peels from the reef edge where bioerosion was noted to
be most extensive. As Text-figure 1 1 shows, more than 40 per cent of the framework has been
destroyed by boring organisms, with Gastrochaenolites being about twice as important as the rest.
Boring bivalves are particularly common at the reef edge; towards the reef core the volume of reef
structure destroyed by borers appears somewhat smaller. We suspect that the amount of material

EXPLANATION OF PLATE 4

Fig. 1 . Natural casts of small borings in the aragonite inner shell layer (now dissolved) of Plicatula , including
Talpina (left centre), Spirichnus (centre and upper right), and Entobia (lower centre), OUM J53680; x 6.
Figs 2-4. Iramena ichnosp., boring of ctenostome bryozoan. 2-3, surface view of boring network in oyster
shell; 2, OUM J53681 ; x 10; 3, OUM J53682; x 16. 4. natural cast in aragonite shell layer (now dissolved)
of Plicatula, OUM J53683; x 15.

PLATE 4

PURSICH et a/., Portland reef borings

150

PALAEONTOLOGY, VOLUME 37

Coombefield

n = 62

Borings in Plicatula

°/o

100 -\
80
60
40

F </ #°v' ^ ^ 0#' o#

y ,<#

ZT

text-fig. 10. Distribution of bor-
ings in Plicatula (percentage of
individuals containing different
borings) based on field observations
on three reefs in Inmosthay Quarry
and one in Coombefield Quarry.

Inmosthay

n = 55

Inmosthay 2

n = 50

Inmosthay 3

actually removed by bioeroders may be even greater than suggested by Text-figure 1 1 because much
of what is now seen as peloidal micrite may in fact originally have been reef framework, now
changed beyond recognition by numerous generations of borers. This has been facilitated by the
rapid synsedimentary cementation of fills of borings which allowed successive generations of borers
to bore into previously excavated, refilled, and cemented borings. We thus do not think it is any
exaggeration to claim the bioerosive processes can be responsible for removal of at least 50 per cent
of the primary framework in some parts of the reef.

Although dwelling-borings were responsible for the great bulk of bioerosion that went on in the
reefs, local evidence of grazing bioerosion is provided by rare bundles of scrape-marks on the inner
faces of a few of the Plicatula (PI. 3, fig. 6). These appear to be confined to the regions of the
adductor muscle scars, either because decaying muscle tissue that attracted small scavengers
lingered here, or perhaps because the myostracal aragonite was softer than the adjacent cross-
lamellar aragonite and thus was more easily scratched.

FURSICH ET AL.: JURASSIC REEFS

151

Volume of framework
removed by bioerosion

4.2%

11.1%

28.7%

points counted
on acetate peels
over 362.3 sq.cm

Small Borings

Bivalve Borings

Organic Frame

Pelleted Micrite
Frame

Others

text-fig. 11. Volumes of reef framework removed by small and large (bivalve) borings in the most heavily
bored areas of the reef (reef edge). Data obtained by point counting on acetate peels from several different reefs.

EVOLUTIONARY AND ECOLOGICAL SIGNIFICANCE OF THE PORTLAND REEFS

Apart from rudist reefs, bivalve-dominated reefs are rare in the Mesozoic and restricted to
comparatively small structures. The ?terquemiid bivalve Placunopsis ostracina Schlotheim built
small domal or cushion-shaped reefs, up to several metres in width, in the Upper Muschelkalk of
the German Basin (e.g. Bachmann 1979; Duringer 1985) and in the Ladinian of Provence,
southeastern France (Brocard and Philip 1989) by cementation of numerous generations on top of
one another, but only after the upper free valves had been removed. Placunopsis is the sole
framebuilder in these reefs; encrusters are rare and consist of Spirorbis , foraminifers, and the bivalve
Enantiostreon (Bachmann 1979). Borings in Placunopsis valves are confined to those of possible
phoronids (Calciroda kraichgoviae Mayer, 1952 and Talpina gruberi Mayer, 1952). Usually they
make up only a few per cent of the reef volume, although in rare cases up to 30 per cent of the
framework volume has been destroyed (Bachmann 1979). The maker of Trypanites , which
frequently occurs on the tops of the domes, postdates reef growth and was not part of the reef
ecosystem.

Bivalve-dominated reefs also occur in the Kimmeridgian of the Lusitanian Basin, Portugal, and
consist of structures, up to 2 m wide and 06-0-8 m high, composed of the oysters Praeexogrya
pustulosa and Nanogyra nana (Fiirsich 1981; Werner 1986). Reef growth was initiated in an
apparently muddy lagoonal environment by overgrowth of large shells (e.g. Isognomon) lying on the
sea floor. Apart from the principal framebuilder ( Praeexogyra pustulosa), the accessory

152

PALAEONTOLOGY, VOLUME 37

framebuilders Dorsoserpula and Juranomia occur in low numbers. Nestlers (the bivalve Arcomytilus)
are rare; vagile elements are more abundant and consist, above all, of a trochid gastropod and the
echinoid Pseudocidaris. Reef destroyers are confined to occasional Gastrochaenolites and Talpina.

Reefs composed largely of bivalves become abundant in the Cretaceous, when various species of
the oyster Crassostrea form, for example, patch reefs in enclosed bays on delta plains on the margin
of the Western Interior Seaway (e.g. Kauffman 1967). Crassostrea is the only framebuilder in these
reefs, associated with a varying number of encrusters (e.g. serpulids) and borers (acrothoracican
barnacles, clionid sponges, and polydorid worms). Nestlers are confined to rare species of
Brachidontes.

All these examples share a very low species diversity. They consist of only one or two
framebuilders, and support a low number of subsidiary encrusters and borers. Their guild structures
are comparable to those of most Mesozoic coral reefs as well as to the Portlandian reefs discussed
here. The main difference is that each guild is occupied by very few members. Part of the reason is
that these bivalve-dominated reefs grew in what appear to have been high stress environments. The
marine fauna of the Muschelkalk is known to be impoverished due to the intermittent isolation of
the German Basin. The Kimmeridgian Praeexogyra-Nanogyra patch reefs apparently grew in
brackish bays and lagoons (Fiirsich and Werner 1986), as did the Cretaceous Crassostrea patch reefs
of the Western Interior Seaway (pers. observations; Fiirsich 1994). The Portlandian patch reefs in
contrast appear to have lived in a more fully marine environment and consequently exhibit a much
higher diversity, though it is interesting to note the low diversity or complete absence of some major
stenotopic groups such as bryozoans, brachiopods, corals, and echinoderms.

There is, however, another feature which distinguishes the Portland reefs not only from other
bivalve-dominated reefs, but also from Jurassic coral reefs growing in comparable settings: the high
diversity and numerical abundance of the reef destroyer guild. The Portland reefs are the oldest
documented case where boring organisms can be shown to destroy such a very large part of the reef
framework and associated cemented interstitial sediment. The diversity and abundance of borers
differs drastically from that of shallow water calcisponge-coral reefs of the Alpine Triassic (e.g.
Fiirsich and Wendt 1977) where bioerosion is insignificant. Shallow water patch reefs of Jurassic
epicontinental seas are bored, to a variable extent, by bivalves ( Gastrochaena , Lithophaga), with few
records of other boring taxa (e.g. Hallam 1975; Fiirsich 1977; Palmer and Fiirsich 1981;
Fagerstrom 1978; Fiirsich and Werner 1991). Although individual coral heads may be quite densely
bored (e.g. Warme and McHuron 1978), the borers are, on the whole, far less conspicuous than in
the Portland reefs. Part of an explanation is surely that extensive hard and solid substrates (shells
of Plicatula and Liostrea; early diagenetically lithified matrix) facilitated boring much more than
porous skeletons of, say, corals and calcareous sponges. Another part of the explanation involves
the excellent preservation of borings facilitated by the early cementation of their fills and subsequent
dissolution of the substrate. However, whereas this aspect probably enhanced the preservation
potential of the various types of small borers (and thus records the original diversity of boring
organisms more faithfully than in many other reefs: examples are Spirichnus and Cunctichnus), it
did not have any influence on the abundance of borers and the percentage of bored reef framework.
Still, these various aspects do not fully account for the importance of borers in the Portland reefs.
It appears that these reefs document an evolutionary radiation of boring organisms that took place
during the Jurassic. Most taxa of borers within the Portland reefs are known from earlier in the
Jurassic or even Triassic (e.g. Talpina , Roger ella , Gastrochaenolites ), but on the whole were less
conspicuous at these earlier times. Bivalve borings have been made by four species within the
Portland reefs in contrast to one or two taxa as recorded from most other Jurassic reefs. In addition,
borings of clionid sponges ( Entobia ), the most abundant element among the smaller borings in the
Portland reefs, are hardly known from older deposits. Published records are vague (e.g. Hallam
1975; Garcia et al. 1989) and may, in fact refer to borings of the haplosclerid sponge Aka which is
not uncommon in the Mesozoic (Reitner and Keupp 1991), or even to other phyla. Apart from the
abundance of borers in general, it is the abundance of Entobia in particular that gives the
Portlandian patch reefs a modern aspect with respect to their reef destroyer guild; clionid sponges

FURSICH ET AL.. JURASSIC REEFS

153

are one of the most important destructive agents of present-day reefal and other hard calcareous
substrates (e.g. Goreau and Hartman 1963; Neumann 1966; Rtitzler 1975; Bromley 1978).

The increasing importance and diversity of the destroyer guild during the Late Jurassic must have
had far-reaching consequences for the reef ecosystem. Not only did it increase substantially the
amount of reef debris, but it could also have promoted framebuilders with fast growth rates such
as colonial scleractinian corals (e.g. Coates and Oliver 1973) within Mesozoic reefs.

SYSTEMATIC PALAEONTOLOGY

This study has recognized several new taxa: one new genus and species of bivalve; one new species
of serpulid; two new ichnogenera and three new ichnospecies of boring. Several first occurrences of
species or higher taxa are recorded for the British Portlandian. In addition, our observations cast
further light on or disagree with some earlier assignments to the Portlandian fauna. These cases are
discussed in this section.

Class bivalvia Linne, 1758
Family arcidae Lamarck, 1809
Genus barbatia Gray, 1842
Subgenus barbatia Gray, 1842

Type species. Area barbata Linne, 1758.

Barbatia ( Barbatia ) bourgueti Loriol, 1892
Plate 5, figure 2

Material. External mould of an articulated specimen, OUM J53657.

Description. Small, articulated shell, less than 10 mm in length, with prominent umbones situated in the
anterior third of shell. Posterior umbonal ridge strong; anteriorly of it there exists a faint, shallow sulcus.
Posterior demarcation of area in form of a thick, rounded rib. Surface ornament consisting of numerous
closely-spaced radial ribs traversed by slightly less conspicuous comarginal growth lines resulting in a reticulate
pattern. Radial ribs posteriorly of umbonal ridge rounded, stronger, and thicker than intervals between them.
Hinge typical of Barbatia.

Remarks. Our specimens belong to a small species nestling in the reef framework. Smaller than de
Loriol’s specimen from the Oxfordian of the Swiss Jura (de Loriol 1892, p. 282, pi. 30, fig. 16), they
exhibit a similar shape and ornamentation. Barbatia kobyi de Loriol (1892, p. 282, pi. 30, figs 17-19)
from the Oxfordian of the Swiss Jura is also similar, but less elongate. Both species come from the
same locality and may well represent different segments of the intraspecific range of a single species.

Family cucullaeidae Stewart, 1930
Genus cucullaea Lamarck, 1801

Type species. Cucullaea auriculifera Lamarck, 1801.

Subgenus idonearca Conrad, 1862

Type species. Cucullaea tippana Conrad, 1858.

Cucullaea ( Idonearca ) sp.

Remarks. The only specimen is an internal mould of a left valve showing a typical Cucullaea hinge.
As the surface ornamentation is not known, a specific designation is not possible. The specimen
resembles in overall shape Cucullaea clathrata Leckenby as figured by Lycett (1863, p. 44, pi. 39,
fig. 4, 4a) from the Bathonian of England.

154

PALAEONTOLOGY, VOLUME 37

Family mytilidae Rafinesque, 1815
Genus lithophaga Roding, 1798
Subgenus lithophaga Roding, 1798

Type species. Lithophaga mytiloides Roding, 1798.

Lithophaga ( Lithophaga ) subcylindrica (Buvignier, 1852)

Plate 5, figures 5-6

Material. Numerous complete internal moulds and, more rarely, external moulds (OUM J53661).

Remarks. Unlined borings of lithophagid bivalves, belonging to the trace fossil Gastrochaenolites
torpedo Kelly and Bromley, 1984, are common in the reef facies (e.g. PL 6, fig. 6). Many of them
contain shells which we refer to Lithophaga (L.) subcylindrica (Buvignier, 1852, p. 22, pi. 17, figs
20-21). This species has been regarded as a junior synonym of L. inclusa (Phillips, 1829) by Pisera
(1987) who argued that Buvignier’s specimens fall within the range of variation of a population of
L. inclusa from the Upper Jurassic of Poland. We feel, however, that a more thorough revision of
the group is needed before morphologically such widely differing taxa are synonymized. The
Portlandian specimens have a straight dorsal and only a gently convex ventral margin compared to
the strongly convex dorsal and ventral margins of typical L. inclusa.

Lithophaga sp. A
Plate 5, figure 8

Material. Two shell fragments (OUM J53662).

Remarks. Lithophaga sp. A differs from all other Lithophaga by displaying distinct radial ribs in the
anterior of the shell which fade from the umbo to the ventral margin and disappear progressively
in a posteroventral direction. The posterior half of one specimen and the posterior third of a second
specimen are smooth except for conspicuous growth lines. Most likely Lithophaga sp. A represents
a new species, but as only two incomplete specimens are available, we prefer an informal designation
until more material becomes available.

Genus modiolus Lamarck, 1 799
Subgenus modiolus Lamarck, 1799

Type species. Mytilus modiolus Linne, 1758.

explanation of plate 5

Figs 1, 9. Gastrochaenopsis recondita (Phillips). 1, OUM J53656; latex cast of external mould of articulated
specimen; dorsal view; x 10. 9, OUM J53663; internal mould of articulated specimen; right valve view; x 7.

Fig. 2. Barbatia (Barbatia) bourgueti de Loriol. OUM J53657; latex cast of external mould of articulated
specimen; dorsal view; x 8.

Figs 3-4, 7. Hiatella ( Pseudosaxicava ) phaseolus (Eudes-Deslongchamps). 3, OUM J53658; internal mould of
articulated specimen; right valve view; x 8. 4, 7, latex casts of external moulds of right valves; 4, OUM
J53659; x8;7, OUM J53660; x 12.

Figs 5-6. Lithophaga ( Lithophaga ) subcylindrica (Buvignier). OUM J53661 ; 5, dorsal view; 6, left valve view
of articulated specimen; x3.

Fig. 8. Lithophaga sp. A. OUM J53662; articulated specimen, right valve view; x 6.

PLATE 5

FURSICH et al Portland reef bivalves

156

PALAEONTOLOGY, VOLUME 37

Modiolus ( Modiolus ) sp.

Material. Fourteen whole and broken specimens (OUM J53707).

Remarks. Small nestling species characterized by an extremely long hinge line, and well rounded
posterior end. Similar in shape to M. jurensis , but the latter is much larger in size.

Family plicatulidae Watson, 1930
Genus plicatula Lamarck, 1801

Type species. Spondvlus plicatus Linne, 1758.

Plicatula damoni Cox, 1937
Plate 3, figure 4

Material. Several hundred internal moulds (OUM J53652).

Remarks. Cox (1929, p. 164, pi. 4, figs 3-4) described specimens of Plicatula from the Late Jurassic
of the Isle of Portland as P. lamellosa n. sp., but later on (Cox 1937) renamed it P. damoni, as
P. lamellosa was preoccupied by a species from the Cretaceous of Russia.

As our material consists of internal moulds only, little can be said about the relationship of
P. damoni to other species of Plicatula. Size ranges up to a length of c. 45 mm; shape is extremely
variable as a result of growth in confined spaces in the reef framework.

Family gastrochaenidae Gray, 1840
Genus gastrochaenopsis Chavan, 1952

Type species. Gastrochaena unicostata Eudes-Deslongchamps, 1838.

Gastrochaenopsis recondite/ (Phillips, 1829)

Plate 5, figures 1, 9

Material. Fifteen specimens, both internal and external moulds (OUM J53656, J53663).

Remarks. Pholas recondita Phillips (1829, pi. 3, fig. 19) was put into the genus Spengleria Tryon,
1862, by Pisera (1987) who regarded Gastrochaenopsis as a junior synonym of the latter. Both

EXPLANATION OF PLATE 6

Figs 1-5, 7. Carterochaena pulcherrima gen. et sp. nov. 1, OUM J53664, paratype; internal mould of
articulated specimen, dorsal view; x 9. 2, OUM J53665, holotype; internal mould of articulated specimen,
dorsal view; x9. 3, OUM J53666, paratype; internal mould of right valve; x7. 4, OUM J53667; internal
mould of articulated specimen in boring, left valve view; x 9. 5, OUM J53688; internal mould of articulated
specimen; dorsal view showing cavities produced by the dissolved shell material of valves and accessory
plate; x 25. 7, OUM J53669; side view of internal mould of anterior region of articulated specimen showing
position of accessory plate; x 25.

Fig. 6. Gastrochaenolites torpedo Kelly and Bromley. OLIM J53670; x 2.

Fig. 8. Caulostrepsis cretacea (Voigt). OUM J53671 ; x 7 5.

Fig. 9. Rogerella pattei (Saint-Seine). OUM J53672; x 6.

Fig. 10. Cunctichnus probans ichnogen. et ichnosp. nov. OUM J53673, holotype; x 2.

PLATE 6

■

; - h.

FURSICH et al Portland reef biota

158

PALAEONTOLOGY, VOLUME 3 7

Gastrochaenopsis and Spengleria are characterized by a triangular area running from the umbo to
the truncated posterior end. This area is separated from the rest of the shell by strongly raised
growth lamellae. In Spengleria , this area is bordered ventrally by a distinct furrow which is absent
from Gastrochaenopsis. The latter genus is therefore kept separate from Spengleria.

The Portland specimens agree well with the material described from the Corallian (Oxfordian) by
Arkell (1933, p. 313, pi. 43, figs 1-4) and from the Oxfordian-Kimmeridgian of Poland by Pisera
(1987, p. 93, pi. 38, figs 1-3, text-figs 7-9). Gastrochaenopsis recondita is more elongate than species
of Gastrochaena but exhibits, like the latter, a distinct anteroventral pedal gape. Internal moulds of
G. recondita do not show a raised area, but have a rounded posteroventrally-directed umbonal
carina followed dorsally by a wide shallow depression. This depression is bordered dorsally by a
second, slightly less conspicuous umbonal carina running to the posterodorsal margin. The elevated
posterior area of the shell, so characteristic of the genus, is therefore exclusively produced by the
strongly raised growth lamellae and is not seen on internal moulds.

On some of the internal moulds, a wide pallial sinus of intermediate depth has been observed.

The boring of the species exhibits a thin calcareous lining which is well preserved and was not,
therefore, originally aragonite. According to its shape, the boring can be accommodated in
Gastrochaenolites dijugus Kelly and Bromley, 1984.

?Family gastrochaenidae Gray, 1840
Genus carterochaena gen. nov.

Derivation of name. After J. G. Carter, from Gastrochaena.

Type species. Carterochaena pulcherrima sp. nov.

Diagnosis. Gastrochaena-\ike bivalve with accessory plate situated anterior of umbones between the
two valves.

Remarks. Carterochaena is identical to Gastrochaena in outline and differs solely by the presence of
the accessory plate. So far, accessory plates are not known in members of the Gastrochaenidae, but
are characteristic of the family Pholadidae. Despite this fact we prefer to group Carterochaena -
with reservation - with the Gastrochaenidae because: (a) none of the other characteristic features
of members of the Pholadidae, such as apophyses, callum, umbonal ventral rib, radial ornament,
etc., are present; and (b) the shape of the valves so strikingly resembles that of Gastrochaena that
without the accessory plate, the specimens would undoubtedly be referred to the latter genus.

Carterochaena pulcherrima sp. nov.

Plate 6, figures 1-5, 7

Derivation of name. Superlative of pulcher (Latin), meaning ‘beautiful’.

Type series. Holotype: OUM J53665 (PI. 6, fig. 2); paratypes: OUM J53664, J53666, J53667 (PI. 6, figs 1, 3,
4 respectively).

Material. I 10 predominantly internal moulds (OUM J53664— J53669).

Diagnosis. Elongate-ovate gastrochaenid with strongly reflected anteroventral margin and elongate
to triangular-ovate accessory plate between the two valves anterior of umbones.

FURSICH ET AL.: JURASSIC REEFS

159

Dimensions of internal moulds (in mm; L, length; H, height).

L

H

L

H

L

H

8-0

4-6

6-4

4-3

6-8

40

1 15

—

60

3-8

8-5

5-4

8-2

5-2

8-0

4-6

7-5

4-5

4-4

2-5

7-5

4-6

6-2

3-8

7-8

4-6

7-5

4-5

4-2

2-8

5-2

31

7-7

4-6

51

3-7

6-5

4-4

7-0

4-7

6-6

4-2

7-5

4-3

6-2

40

6-7

3-9

8-4

5-0

7-5

41

8-8

4-4

61

3-0

61

3-9

Description. Small (on average 5-10 mm long), moderately inflated shell. Umbones prosogyrate, situated close
to the anterior end. Dorsal margin straight, grading into well rounded posterior margin. Ventral margin faintly
convex posteriorly, but strongly reflected anteriorly resulting in a straight to faintly concave anterior margin
and a wide antero ventral pedal gape. Anterodorsal shell margin slightly reflected along the plane of
commissure. Anterior adductor muscle scar situated in the anterodorsal corner of the valve; posterior adductor
scar somewhat larger. Pallial sinus wide and deep. Shell ornamented with numerous fine, faintly rugose growth
lines.

Accessory plate elongate ovate to rounded triangular in shape with a beak-like end pointing in the direction
of the umbones. The accessory plate closely fits between the upturned anterodorsal shell margins. Judging from
the cavity space between internal and external moulds (PI. 6, fig. 5), the thickness of the plate was at least twice
that of the valves.

Remarks. The main diagnostic feature of Carterochaena pulcherrima is the presence of a thick
accessory plate. In the Pholadidae such plates, when situated anteriorly of the umbones (e.g.
protoplax), serve to protect the anterior adductor muscle which inserts on the reflected dorsal
margin. The anterior adductor muscle inserts very closely to the anterodorsal margin and transverses
right behind the pedal gape in pholadids and C. pulcherrima , and a similar function of the accessory
plate seems likely. However, the facts that the plate is much thicker than the rest of the shell and
that it is positioned at the anterodorsal end of shell and projecting slightly beyond it, suggest that
it may well have aided the boring process by mechanically abrading hard substrates that had been
chemically weakened. Any shell material that was lost during this process could have been replaced
by addition of new layers on the inside of the plate.

C. pulcherrima occurs in Gastrochaenolites dijugus Kelly and Bromley, 1984, flask-shaped borings
characterized by a restricted neck region with a figure of eight cross-section. The boring exhibits a
thin calcareous lining which, as in the case of G. recondita , was not aragonitic.

Genus hiatella Bose, 1801

Type species. Hiatella monoperta Bose, 1801.

Subgenus pseudosaxicava Chavan, 1952
Type species. Pseudosaxicava bernardi Chavan, 1952.

Hiatella ( Pseudosaxicava ) phaseolus (Eudes-Deslongchamps, 1838)

Plate 5, figures 3-4, 7

Material. Sixty five specimens, both internal and external moulds and casts (OUM J53658-J53660).

160

PALAEONTOLOGY, VOLUME 37

Remarks. The Portlandian specimens were described as ‘Area’ foetida n. sp. by Cox (1929,
p. 140, pi. 1, figs 2-3). Kelly (1980) placed A. ' foetida with Hiatel/a ( Pseudosaxicava ) and, later
(Kelly 1985, p. 172, fig. 1), regarded it as a junior synonym of Hiatella ( Pseudosaxicava ) arcadiformis
(Keeping, 1883). Pisera (1987, p. 90, pi. 37, figs 1-4, text-figs 5-6) regarded H. ( Pseudosaxicava )
arcadiformis as a junior synonym of Saxicava phaseolus Eudes-Deslongchamps (1838, p. 227, pi. 9,
figs 25-26, 33—34), arguing that the different species represent only intraspecific variants of a
morphologically very plastic species. Pisera’s view is followed here.

This highly variable species, up to 15 mm in length (average 6-8 mm), is characterized by a
subrectangular outline and a strong posterior carina running to the posteroventral shell margin.
A second, less conspicuous carina runs from the umbo to the posterodorsal margin. Silicon rubber
casts of external moulds reveal faint growth lines and an ornament of radially arranged lines of
minute pustules. This ornament extends across most of the shell (PI. 5, fig. 7) or else is restricted to
patches on the flank of the valves (e.g. PI. 5, fig. 4). Such pustules may correspond to periostracal
spikes (see also Pisera 1987). Clavulate spines also occur at regularly spaced intervals on both
carinae. Some internal moulds exhibit large adductor muscle scars and a wide, moderately deep
pallial sinus (PI. 5, fig. 3).

The great variability of the species is largely a result of its nestling habit. Hiatella ( Pseudosaxicava )
phaseolus was found, above all, in Gastrochaenolites torpedo (which, in addition, often contained
valves of the borer Lithophaga) and in G. dijugus.

Phylum annelida Lamarck, 1809
Family serpulidae Burmeister, 1837
Genus serpula Linne, 1758

Type species. Tubus vermicularis Ellis, 1755.

Subgenus cycloserpula Parsch, 1956
Type species. Serpula flaccida Goldfuss, 1831.

Serpula ( Cycloserpula ) striatissima sp. nov.

Plate 3, figure 3

Derivation of name. Superlative of striatus (Latin), meaning ‘striped’.

Type series. Holotype: OUM J53653.

Material. Numerous external moulds embedded into the surface of internal moulds of Plicatula.

Diagnosis. Serpula with round cross-section and seven axial riblets.

Description. Small, encrusting serpulid, 0-3-0- 7 mm in diameter, with irregular, winding course. Usually, the
tubes disappear into the substrate after having meandered across the surface for most of their length which
suggests that the youngermost parts of the tubes became detached from the substrate and grew away from it.

Cross-section round. Surface ornament of seven axial riblets crossed by numerous fine growth lines,
producing a slightly reticulate pattern, and more widely spaced annulae.

Remarks. Despite the large number of serpulid species known from the Jurassic the present
specimens do not match any of them. They are therefore accommodated in the new species
striatissima within the form subgenus Cycloserpula.

FURSICH ET AL.\ JURASSIC REEFS

161

ICHNOTAXA

Ichnogenus arachnostega Bertling, 1992
Type ichnospecies. Arachnostega gastrochaenae Bertling, 1992.

Arachnostega gastrochaenae Bertling, 1992
Plate 3, figures 1-2, 4

Material. Numerous specimens in fills of bivalves (OUM J53650-J53652).

Description. Irregularly ramifying and anastomosing burrows with circular to oval cross-section. Preserved as
negative reliefs or calcite-filled endoreliefs at the surface of internal moulds of Plicatula damoni or other
bivalves such as Lithophaga. The burrows are roughly cylindrical, branch frequently but at irregular intervals,
whereby branches join up to form networks. Several ‘ontogenetic’ stages are present: the tunnels of the
smallest networks are around 0 03 mm in diameter, those of the largest networks 1-5 mm. Average tunnel
diameter is 0-3-0-5 mm. The diameter of individual networks remains fairly constant, but networks of different
sizes are connected with each other. Also, parts of smaller networks are commonly seen to branch off (or join?)
larger ones.

Remarks. The burrows closely resemble those recently described from internal moulds of molluscs
from Late Jurassic reefs of northern Germany (Bertling 1992). Similar traces were recorded by
Reineck (1980) from the tidal flats of the North Sea, there made by the polychaetes Heteromastus
and Nereis in stiff sediment fills of the bivalve Mya arenaria. Accordingly, Bertling ( 1992) assumed
errant polychaetes to be the producers of the Jurassic burrow systems.

A. gastrochaenae is a ubiquitous trace in internal moulds of bivalves in the Portlandian reefs. Very
likely, their producers preferred internal sediments as these might have contained a higher
concentration of bacteria and other organic matter produced by the decay of the original soft parts.
Moreover the Recent Heteromastus filiformis is known to live in oxygen-poor sediments (e.g. Dorjes
1978, p. 134) and thus is adapted to exploit the niche of sediment-filled shells. However, it is unlikely
that the producers of A. gastrochaenae were confined to the environments from which the trace is
known so far. It is the low preservation potential of such traces outside the shell-protected
microhabitat, especially their disappearance during compaction, that creates such a specific
distribution pattern.

As Bertling (1992) suggested, the different-sized meshworks most likely represent different age
groups within the population of the producer. Such burrows are built in such a short time that no
size increase due to growth of the producer can be usually detected.

Ichnogenus caulostrepsis Clarke, 1908
Type ichnospecies. Caulostrepsis taeniola Clarke, 1908.

Caulostrepsis cretacea (Voigt, 1971)

Plate 6, figure 8

Material. More than twenty five natural casts in dissolved Plicatula shells (OUM J53671).

Description. Pouch-like flattened boring, 5-12 mm in length with a flat-oval cross-section (for terminology see
Bromley and d’Alessandro 1983, fig. 2). Near point of origination, the width-height ratio is about two and

162

PALAEONTOLOGY, VOLUME 37

increases up to four near the vertex. Course of boring straight to sometimes sinuous or twisted in a step-like
fashion. No vane developed, but an axial depression is commonly present.

Remarks. Bromley and d'Alessandro (1983) discussed several ichnospecies of Caulostrepsis in detail.
Our material fits best the description of C. cretacea , although an axial depression (not corresponding
to a vane) is far more common and pronounced than in the material described by Bromley and
d’Alessandro (1983). C. taenida Clarke, 1908 differs by possessing a distinct vane and a
corresponding dumb-bell cross-section.

The producer of the borings was most likely a polychaete (compare Recent borings of Polydora
and Dodecaceria).

Ichnogenus cunctichnus ichnogen. nov.

Derivation of name, cunctare (Latin), meaning ‘hesitate’ (after the behaviour pattern of the trace fossil
producer); ichnos (Greek), meaning ‘trace’.

Type ichnospecies. Cunctichnus probans ichnosp. nov.

Diagnosis. Cylindrical borings in shells, arcuate to highly sinuous or planispiral, with thin, short,
tapering side-branches at points where the tubes abruptly change direction.

Cunctichnus probans ichnosp. nov.

Plate 6, figure 10; Text-figure 12a
Derivation of name. (Latin) probare , meaning ‘to probe’.

Type series. Holotype: OUM J53673.

Material. More than 20 specimens occurring as natural casts in dissolved Plicatula shells.

Diagnosis. As for ichnogenus.

Description. Cylindrical tunnels of a constant diameter ranging from 0-8 to 4 mm in different specimens. Cross-
section circular to slightly elliptical in the plane of boring. Borings occur as casts of tunnels in dissolved shells
of Plicatula damoni. Course of tunnels arcuate, planispiral, or highly sinuous, defined by shape of available
substrate. Where the arcuate tunnel casts approach the margin of the host substrate, thin and short side-
branches with a tapering or blunt end, often not more than 1-2 mm in length and less than 1 mm in diameter,
extend either at some angle from or in continuation of the main shaft. The main shaft then continues at an
angle (sometimes more than 90°) from the original direction until the margin of the host substrate is
encountered somewhere else. The casts never cross each other, very rarely touch each other, but often run
parallel to each other with less than 01 mm distance between them. The cross-section of the short side-branches
varies from circular to distinctly flattened. Often, two or more side-branches occur side by side.

Elongate-oval bulges. 0 1—0-2 mm across, arranged parallel to the long axis of the cylinder, were observed
m one specimen. They correspond to shallow scoops on the walls of the boring and most likely represent
individual phases of the excavation process.

Remarks. There can be little doubt that the course of the boring is determined to a large extent by
the shape of the substrate. The side-branches represent probing tunnels through which the organism
was able to sense the edge of the substrate. This, as well as the fact that the borings do not cut
themselves but may run closely parallel to an earlier limb, points to a negative thigmotactic

FURSICH ET AL.. JURASSIC REEFS

163

text-fig. 12. Sketches of branching patterns in new ichnospecies of borings from the Portland reefs.
a. Cunctichnus probans ichnogen. et ichnosp. nov. showing thin and short side branches where animal has
approached the outer edge of the Plicatula substrate, has withdrawn a couple of millimetres, and has set off
again in a different direction in order to remain within the shell; scale-bar is 5 mm. b, branching in Talpina

bromleyi ichnosp. nov.; scale-bar is 2 mm.

behaviour of the producer. The thin side-branches suggest that the first step in boring was a short,
thin exploratory shaft which, when the substrate proved satisfactory, was enlarged to its final size.
This behaviour minimized energy expenditure in case the adopted direction had to be aborted. The
probing tunnels indicate that the producer was an animal with a proboscis. A sipunculid origin of
the boring is therefore likely.

Possibly a somewhat similar behaviour is displayed by some Vermiforichnus clarkei Cameron,
1969, from the Devonian Hamilton Group of New York State when these tunnels reach the limits
of their brachiopod host substrates (e.g. Hiintzschel 1975, fig. 82-lu).

Ichnogenus entobia Bronn, 1838
Type ichnospecies. Entobia cretacea Portlock, 1843.

Entobia cervicornis ichnosp. nov.

Plate 7, figures 1-2, 4

Derivation of name. (Latin) cervus , meaning ‘deer’; (Latin) cornua, meaning ‘antlers'.

Type series. Holotype: OUM J53674 (PI. 7, fig. 1); paratypes: OUM J53675— J53676 (PI. 7, figs 2, 4).

Material. Numerous specimens, preserved as natural casts in dissolved Plicatula shells.

Diagnosis. Non-camerate Entobia forming an irregular boxwork in growth phase C (sensu Bromley
and d’Alessandro 1984). Phase A represented by long slender tubules grading into thicker, antler-
shaped cylindrical tubes of growth phase B. Phases D E not developed. Typical chip micro-
ornament, as seen in most clionid borings, is absent.

Description. Anastomosing Entobia of widely differing diameter. Long, slender tubules (exploratory threads)
of phase A (for terminology see Bromley and d’Alessandro 1984) widespread, around 50 //m in diameter

164

PALAEONTOLOGY, VOLUME 37

(PI. 7, fig. 2). Growth phase B is also very common and represented by antler-shaped cylindrical tubes,
100-300 pm in diameter, branching in three dimensions. The final developmental stage is phase C in which
galleries, subcylindrical to oval in cross-section, connect to form an irregular anastomosing boxwork. At
branching points, several branches often take off, usually parallel to laminae of the microstructure of the host
shell. Diameter of galleries varies between 350 and 850 //m. Apertures rarely observed, around 250 pm in
diameter.

Remarks. In shape and branching pattern, this ubiquitous boring is similar to Entobia megastoma
(Fischer, 1868) as described by Bromley and d1 Alessandro (1984). Major differences are the
distinctly smaller size of E. cervicornis and the lack of developmental stage D. Thus, the boxwork
of E. cervicornis never reaches the stage where most of the substrate has been removed and only thin
columns are left between the galleries as is the case in E. megastoma. The oval cross-section of some
phase C galleries probably results from the parallel arrangement of these galleries to the laminae of
the shell microstructure.

Like other species of Entobia , E. cervicornis can be attributed to the boring activity of clionid
sponges.

Ichnogenus iramena Boekschoten, 1970
Type ichnospecies. Iramena danica Boekschoten, 1970.

Iramena ichnosp.

Plate 4, figures 2^4

Material. Several colonies in two shells of Liostrea expansa. Three specimens preserved as natural internal
moulds (OUM J53681-J53683).

Remarks. The ichnogenus Iramena was erected by Boekschoten (1970, p. 45) for bryozoan borings
with round to reniform apertures of zooids and an irregular tunnel network representing stolons.

Borings of bryozoans are claimed by both bryozoologists and ichnologists. As no parts of the
producing bryozoans are left after their decay, the boring has to be treated as a trace fossil, although
it apparently accurately reflects the shape of the organism (see discussion in Hantzschel 1975,
p. W136; cf. Pohowsky 1974).

Iramena is relatively rare in the Portland reefs. The short, vertical to oblique tunnels (zooids),
which are arranged at right angles to, and alternately along the tunnel network (stolons), are
connected to it by thin tunnels (peduncles). Depending on the degree of erosion of the shell surface,
the zooidal tunnels are round to tear-shaped in cross-section. Their apertures are circular and
measure 30-35 /mi in diameter. The zooidal tunnels extend, slightly inclined to faintly curved, about
100-150 //m into the substrate. Their cross-section is oval and they exhibit a tapering end.

The shape of the borings corresponds well to that of the ctenostome bryozoan genus Penetrantia
Silen, 1946, which is characterized by pedunculate zooids orientated more or less vertically in the

EXPLANATION OF PLATE 7

Figs 1-2, 4. Entobia cervicornis ichnosp. nov. 1. OUM J53674, holotype; growth phases B and C; x 5. 2, OUM
J53675, paratype; growth phases A and B; x 5. 4, OUM J53676, paratype; growth phase B; x 5.

Figs 3, 6. Spirichnus spiralis ichnogen. et ichnosp. nov. 3, OUM J53677, paratype; x 6. 6, OUM J53678,
holotype; x 6.

Fig. 5. Talpina bromleyi ichnosp. nov. OUM J53679, holotype; x 7-5.

PLATE 7

FURSICH el al., Entobici , Spirichnus , Talpina

166

PALAEONTOLOGY, VOLUME 37

substrate (Pohowsky 1978). The genus was known previously only from the mid-Cretaceous to the
Recent. Its occurrence in the Portland reefs extends its record back to the Late Jurassic.

The internal moulds represent a new species of Penetrantia , differing distinctly from those
described so far.

Ichnogenus rogerella Saint-Saine, 1951
Type ichnospecies. Rogerella lecontrei Saint-Saine, 1951.

Rogerella pattei (Saint-Seine, 1954)

Plate 6, figure 9

Material. Numerous natural casts in dissolved Plicatula shells (OUM J53672).

Description. Elongate to short, pouch-shaped borings with expanded and flattened anterior. Maximum width
1-5—1 *7 mm; length 2-7-3 mm.

Remarks. Several ichnogenera such as Rogerella Saint-Seine, 1951, Simonizapfes Codez and Saint-
Seine, 1958, and Zapfella Saint-Seine, 1954 exist for sac- to pouch-shaped borings with slit-like
openings which are made by acrothoracican barnacles. We agree with Bromley and d’Alessandro
(1987) that one ichnogenus is sufficient to accommodate most cirriped borings. The Portland
material can therefore be referred to the earliest-named of these available ichnogenera, Rogerella.
There is no significant difference between our material and Rogerella pattei (Saint-Seine, 1954,
p. 448) to which it is therefore referred.

On Plicatula damoni , R. pattei usually occurs in clusters of three to twelve individuals.

Ichnogenus spirichnus ichnogen. nov.

Derivation of name. (Latin) spira, meaning ‘anything coiled’; (Greek) ichnos, meaning ‘trace’.
Type ichnospecies. Spirichnus spiralis ichnosp. nov.

Diagnosis. Spirally-coiled cylindrical spiral borings.

Spirichnus spiralis ichnosp. nov.

Plate 7, figures 3, 6

Derivation of name. (Latin) spiralis, meaning ‘spiral’.

Type series. Holotype: OUM J53678 (PI. 7, fig. 6); paratype: OUM J53677 (PI. 7, fig. 3).

Material. Numerous specimens and fragments of varying length.

Diagnosis. Cylindrical, irregularly spirally-coiled borings of 0-5 mm in diameter, branching at highly
irregular intervals.

Description. Cylindrical casts of constant diameter (0-5 mm) which are spirally coiled. Coils usually tight, but
looser at irregular intervals. Where the spiral coils are closely set, the cross-section of the casts is not strictly
circular, but appears to have a bulge towards the coiling axis. Several specimens therefore show a vane-like
development.

FURSICH ET AL.. JURASSIC REEFS

167

Branching occurs commonly, but at highly irregular intervals. Unbranched cylinders up to 23 mm in length
have been observed, but also specimens where Y-shaped branching occurred every 2-3 mm. The branching
does not occur in a plane, but produces a three-dimensional network. Density of Spirichnus spiralis is very high
in some specimens resulting in a dense boxwork of cylindrical casts. In other specimens, S. spiralis appear to
be confined to certain shell layers and the resulting network forms a plane with the spiral axes running parallel
to it. In these cases, the spiral coils are often compressed laterally, apparently due to lack of usable space.

Remarks. No comparable borings are known from the literature. Spirichnus spiralis resembles
Meandropolydora barocca Bromley and d'Alessandro (1987, p. 386, pi. 40, figs 1-3; pi. 43, fig. 3;
pi. 44, figs 1-2, 5; pi. 47, fig. 1 ; pi. 48, fig. 4; text-figs 12-14) from the Pleistocene of southern Italy
in cases where the bulged circular cross-section hints at the presence of a vane. However, pouches
are never developed and S. spiralis must therefore be kept as a separate ichnotaxon. Its producer
remains unknown, but probably belongs to one of the several phyla of ‘worms’.

Ichnogenus talpina Hagenow, 1840
Type ichnospecies. Talpina ramosa Hagenow, 1840.

Talpina bromleyi ichnosp. nov.

Plate 7, figure 5; Text-figure 12b
Derivation of name. After R. G. Bromley.

Type series. Holotype: OUM J53679 (PI. 7, fig. 5).

Material. Several hundred specimens on internal moulds of Plicatula.

Diagnosis. Straight to gently curved branched cylindrical borings, 0 25 mm in diameter. Distance
between branching points much longer than in other species of Talpina.

Description. Straight to gently curved cylindrical casts, running parallel and close to the surface of the original
shell. Diameter of casts 0-25 mm. Branching occurs, but is far less common than in other ichnospecies of
Talpina. Branches usually take off from the convexly curved sides of borings. No anastomosing was observed.
Where density of boring is low, the galleries run close to the surface of the shell; in densely bored shells, the
galleries extend to lower levels, 1-2 mm below the surface. Rarely, shallow, simple or combined U-tubes from
which horizontal galleries branch off connect to the surface. Where density of boring is very high, frequently
two or three galleries can be observed to run parallel to each other for several millimetres, separated from each
other by only a few microns, but never touching. T. bromleyi never cut each other, but cross over or under an
already existing tunnel.

Remarks. Diameter of borings, branching pattern, and arrangement of galleries with respect to the
surface are typical of Talpina. The very similar ichnogenus Conchotrema differs by having a much
more irregular network and by displaying anastomosis. T. bromleyi differs from other ichnospecies
of Talpina (T. ramosa Hagenow, 1840, T. hirsuta Voigt, 1975, T. eduliformis Quenstedt, 1858,
T. gruberi Mayer, 1952, and T. anmdata Voigt, 1975) by displaying a markedly lower frequency of
branching. The relatively great depth of penetration seen in densely bored Plicatula shells is
obviously a result of crowding. The apertures of the lowest networks may possibly have connected
to abandoned shallower galleries and not necessarily straight to the surface.

Acknowledgements. Our grateful thanks go to all the staff at ARC (South Western), particularly to Mr
E. Maidment and Mr T. L. Poole for much help and interest over the last six years. A. Satterley and R. Clother

168

PALAEONTOLOGY, VOLUME 37

helped with the field work; H. Schirm took the photographs. T.J.P. and K.L.G. gratefully acknowledge
receipt of an Undergraduate Bursary from the Nuffield Foundation.

REFERENCES

arkell, w. J. 1933. The Jurassic System in Great Britain. Oxford University Press, London, 681 pp.

1947. Geology of the country around Weymouth, Swanage, Corfe and Lulworth. Memoirs of the
Geological Survey of Great Britain ( Explanation of Sheets 34 1, 342, 343, with small portions of Sheets 327,
328, 329), 1-386.

bachmann, G. h. 1979. Bioherme der Muschel Placunopsis ostracina v. Schlotheim und ihre Diagenese. Neues
Jahrbuch fur Geologie und Palaontologie, Abhandlungen, 158, 381-407.
bertling, m. 1992. Arachnostega n. ichnog. - burrowing traces in internal moulds of boring bivalves (Late
Jurassic, Northern Germany). Palaontologische Zeitschrift , 66, 177-185.
blare, j. F. 1880. On the Portland rocks of England. Quarterly Journal of the Geological Society, London, 36,
189-236.

boekschoten, g. j. 1970. On bryozoan borings from the Danian at Fakse, Denmark. 43-48. In crimes, t. p.

and harper, j. c. (eds). Trace fossils. Geological Journal Special Issue, 3, 547 pp.
bosc, l. a. g. 1801 . Histoire naturelle des vers, contenant lew description et lew moeurs, 1. Roret, Paris, 150 pp.
brocard, c. and philip, J. 1989. Les bioconstructions a Placunopsis ostracina v. schlotheim dans le Ladinien
superieur de Provence (SE de la France). Bulletin de la Societe Geologique de France, (8), 5, 1201-1206.
bromley, r. G. 1978. Bioerosion of Bermuda reefs. Palaeogeographv, Palaeoclimatologv, Palaeoecologv, 23,
169-197.

-and d’alessandro, a. 1983. Bioerosion in the Pleistocene of southern Italy: ichnogenera Caulostrepsis
and Meandropolydora. Rivista Italiana di Paleontologia e Stratigrafia, 89, 283-309.

1984. The ichnogenus Entobia from the Miocene, Pliocene and Pleistocene of southern Italy. Rivista
Italiana di Paleontologia e Stratigrafia, 90, 227-296.

1987. Bioerosion of the Plio-Pleistocene transgression of Southern Italy. Rivista Italiana di
Paleontologia e Stratigrafia, 93, 379-442.

bronn, H. G. 1838. Lethaea Geognostica oder Abbildungen und Beschreibungen der fur die Gebirgsformationen
bezeichnendsten Versteinerungen, 2, 673-1350. Schweizerbart, Stuttgart.
buckland, w. and de la beche, h. t. 1836. On the geology of the neighbourhood of Weymouth and the
adjacent parts of the coast of Dorset. Transactions of the Geological Society, London, (2), 4, 1^16.
burmeister, c. H. c. 1837. Handbuch der Naturgeschichte. Berlin, 858 pp.

buvignier, a. 1852. Statistique geologique, miner alogique , mineralurgique et paleontologique du departement de
la Meuse. Atlas. Bailliere, Paris, 52 pp.

carter, j. g. 1978. Ecology and evolution of the Gastrochaenacea (Mollusca, Bivalvia) with notes on the
evolution of the endolithic habitat. Peabody Museum of Natural History, Yale University, Bulletin, 41, 1-92.
cameron, b. 1969. Paleozoic shell-boring annelids and their trace fossils. American Zoologist, 9, 689-703.
chafetz, H. s. 1986. Marine peloids: a product of bacterially induced precipitation of calcite. Journal of
Sedimentary Petrology, 56, 812-817.

chavan, a. 1952. Les pelecypodes des sables astartiens de Cordebugle (Calvados). Memoires Suisses de
Paleontologie, 69, 1-132.

Clarke, J. m. 1908. The beginnings of dependent life. Bulletin of the New York State Museum, 121, 146-169.
coates, a. G. and Oliver, w. a., jr 1973. Coloniality in zoantharian corals. 3-27. In boardman, r. s.,
CHEETHAM, a. H. and Oliver, w. A., jr (eds). Animal colonies. Development and function through time.
Dowden, Hutchinson & Ross, Stroudsburg, 604 pp.

codez, j. and saint-saine, r. 1958. Revision des cirripedes acrothoraciques fossiles. Bulletin de la Societe
Geologique de France, (6), 7, 699-719.

conrad, t. a. 1858. Descriptions of two new genera of shells. Proceedings of the Academy of Natural Sciences
of Philadelphia, 11,1 65- 1 66.

1 862. Descriptions of new genera, subgenera and species of Tertiary and Recent shells. Proceedings of the
Academy of Natural Sciences of Philadelphia, 1862, 284-291.

COPE, J. C. W., DUFF, K. L., PARSONS, C. F., TORRENS, H. S., WIMBLEDON, W. A. and WRIGHT, J. K. 1980. A
correlation of Jurassic rocks in the British Isles. Part Two; Middle and Upper Jurassic. Geological Society,
London , Special Report, 15, 1-109.

cox, l. r. 1929. A synopsis of the Lamellibranchia and Gastropoda of the Portland Beds of England. Part I.
Lamellibranchia. Proceedings of the Dorset Natural History and Archaeological Society, 50, 131-202.

FURSICH ET AL.\ JURASSIC REEFS

169

— 1937. Notes on Jurassic Lamellibranchia. IV. A re-named British Portlandian species. Proceedings of the
Malacological Society , London , 22, 204.

damon, R. 1884. Geology of Weymouth , Portland , and coast of Dorsetshire, from Swanage to Bridport-on-the-
Sea. 2nd edition. Edward Stanford, London, 250 pp.

dorjes. j. 1978. Das Watt als Lebensraum. 107-143. In reineck, h.-e. (ed. ). Das Watt. Waldemar Kramer,
Frankfurt, 142 pp.

duringer, p. 1985. Strategic adaptive de la croissance de Placunopsis ostracina Schlotheim, epizoaire du
Muschelkalk superieur (Trias germanique, Est de la France). Neues Jahrbuch fur Geologie und Paldontologie,
Monatshefte , 1985, 1-22.

ellis, J. 1755. An essay towards a natural history of the corallines and other marine productions of the like kind ,
commonly found on the coasts of Great Britain and Ireland. London, xvii -F 103 pp.
eudes-deslongchamps, j. a. 1838. Memoire sur les coquilles fossiles lithophages des terrains secondaires du
Calvados. Memoires de la Societe Linneenne de Normandie , 6, 220-229.
fagerstrom, j. a. 1987. The evolution of reef communities. John Wiley and Sons, New York, 599 pp.
fischer. m. p. 1868. Recherches sur les eponges perforantes fossiles. Nouvelles Archives du Museum d'Histoire
Naturelle , Paris , 4. 1 17-173.

fitton, w. h. 1836. Observations on some of the strata between the Chalk and the Oxford Oolite in the south-
east of England. Transactions of the Geological Society , London , (2), 4, 103-383.
francis, j. E. 1984. The seasonal environment of the Purbeck (Upper Jurassic) fossil forest. Palaeogeography ,
Palaeoclimatology , Palaeoecology , 48, 285-307.

— 1986. The calcareous paleosols of the basal Purbeck Formation (Upper Jurassic), southern England.
112-138. In wright, v. p. (ed.). Paleosols - their recognition and interpretation. Blackwells, Oxford, 186 pp.

fursich, f. t. 1977. Corallian (Upper Jurassic) marine benthic associations from England and Normandy.
Palaeontology , 20, 337-385.

— 1981. Salinity-controlled benthic associations from the Upper Jurassic of Portugal. Lethaia. 14, 203-223.

— 1994. Palaeoecology and evolution of Mesozoic salinity-controlled benthic macroinvertebrate associa-
tions. Lethaia , 26, 327-346.

— and wendt, j. 1977. Biostratinomy and palaeoecology of the Cassian Formation (Tnassic) of the
Southern Alps. Palaeogeography , Palaeoclimatology, Palaeoecology , 22. 257-323.

— and werner, w. 1986. Benthic associations and their environmental significance in the Lusitanian Basin
(Upper Jurassic, Portugal). Neues Jahrbuch fid Geologie und Paldontologie , Abhandlungen , 172, 271-329.

1991. Palaeoecology of coralline sponge-coral meadows from the Upper Jurassic of Portugal.
Paldontologische Zeitschrift, 65, 35-69.

garcia. j.-p., courville, p., laurin, b. and thierry, ). 1989. Degradation differentielle et encroutement des
constructions a madreporaires du Callovien inferieur (Jurassique moyen) d'Etrochey (Cote-d'Or). Bulletin
de la Societe Geologique de France , (8), 6, 1217-1225.
goldfuss, A. 1831. Petrefacta Germaniae. 1, 3. Lieferung. Diisseldorf, 224-242.

goreau, t. f. and hartman, w. d. 1963. Boring sponges as controlling factors in the formation and
maintenance of coral reefs. 25-54. In soggnaes, r. f. (ed.). Mechanisms of hard tissue destruction.
Publications , American Association for the Advancement of Science , 75, 1-264.
gray, j. e. 1840. Synopsis of the contents of the British Museum. 42nd edition. British Museum, London,
370 pp.

— 1842. Synopsis of the contents of the British Museum. 44th edition. British Museum, London, 308 pp.
hantzschel, w. 1975. Trace fossils and Problematica. In moore, r. c. (ed.). Treatise on invertebrate

paleontology. Part W. Miscellanea , Supplement I. Geological Society of America and University of Kansas
Press, Boulder, Colorado and Lawrence, Kansas, 269 pp.
hagenow, F. von 1840. Monographie der Riigen'schen Kreide-Versteinerungen. II. Abteilung. Radiarien und
Annulaten. Neues Jahrbuch fur Mineralogie , Geognosie , Geologie und Petrefaktenkunde, 1840, 631-672.
hallam, a. 1975. Coral patch reefs in the Bajocian (Middle Jurassic) of Lorraine. Geological Magazine , 112,
383-392.

hinde, G. j. 1887-1912. A monograph of the British fossil sponges. I. Sponges of Palaeozoic and Jurassic strata.
Monograph of the Palaeontographical Society, 40 (190) [1887; pp. 1-92, pis 1-8], 41 (194) [1888; pp. 93- 188;
pi. 9], 47 (222) [1893; pp. 189-254, pis 10-19], 65 (322) [pp. 255-264],
kauffman, E. G. 1967. Coloradoan macroinvertebrate assemblages, central Western Interior, United States.
67-143. In kauffman, e. g. and KENT, H. c. (eds). Paleoenvironments of the Cretaceous Seaway in the Western
Interior - a symposium. Colorado School of Mines, Golden, 224 pp.

170

PALAEONTOLOGY, VOLUME 37

keeping, w. 1883. The fossils and palaeontological affinities of the Neocomian deposits ofUpware, and Brickhill.
University Press, Cambridge, 167 pp.

kelly, s. r. a. 1980. Hiatella - a Jurassic bivalve squatter? Palaeontology , 23, 769-781.

1985. On the identity of the Upper Jurassic, Portland Beds bivalve ‘Area ’ foetida Cox, 1929. Proceedings
of the Dorset Natural History and Archaeological Society, 106, 172.

and bromley, r. G. 1984. Ichnological nomenclature of clavate borings. Palaeontology , 27, 793-807.
kidwell, s. m. and jablonski, d. 1983. Taphonomic feedback. Ecological consequences of shell accumulation.
195-248. In tevesz, m. j. s. and McCall, p. l. (eds). Biotic interactions in Recent and fossil benthic
communities. Plenum Press, New York, 837 pp.

lamarck, j. b. de 1799. Prodrome d’une nouvelle classification des coquilles comprenant une redaction
appropriee des caracteres generiques, et l'etablissement d’un grand nornbre de genres nouveaux. Memoires
de la Societe d'Histoire Naturelle Paris, 1, 63-91.

- 1801. Systeme des animaux sans vertebres. Paris, viii + 432 pp.

— 1809. Plulosophie zoologique etc. 2 vols, Paris, 422 and 473 pp.

linne, c. von 1758. Systema naturae per regna tria naturae secundum classes , ordines, genera, species, cum
char act eribus, differenciis, synonymis, locis. 10th edition, 1. Stockholm, 1-824.
loriol, p. de 1892. Etudes sur les mollusques des couches coralligenes mferieures du Jura Bernois, 4e partie.

Memoires de la Societe Paleontologique de la Suisse, 19, 259-419.
lycett, j. 1863. Supplementary monograph on the Mollusca from the Stonesfield Slate, Great Oolite, Forest
Marble, and Cornbrash. Monograph of the Palaeontographical Society, 15 (65), 1-129, pis 31-45.
macintyre, i. G. 1985. Submarine cements; the peloidal question. 109-1 16. In schneidermann, n. and Harris,
p. M. (eds). Carbonate cements. Society of Economic Paleontologists and Mineralogists, Special Publication,
36, 379 pp.

mayer, g. 1952. Lebensspuren von Bohrorganismen aus dem unteren Hauptmuschelkalk (Trochitenkalk) des
Kraichgaues. Neues Jahrbuch fur Geologie und Paldontologie, Monatshefte, 1952, 450 — 456.
mock, s. e. and palmer, t. j. 1991. Preservation of siliceous sponges in the Jurassic of southern England and
northern France. Journal of the Geological Society, London , 148, 681-689.
neumann, a. c. 1966. Observations on coastal erosion in Bermuda and measurements of the boring rate of the
sponge, Cliona lampa. Limnology and Oceanography, 11, 92-108.
oschmann, w. 1988. Upper Kimmeridgian and Portlandian marine macrobenthic associations from southern
England and northern France. Facies , 18, 49-82.

palmer, T. j. and fursich, F. T. 1981. Ecology of Bathonian sponge reefs from Normandy. Palaeontology, 24,
1-23.

parsch, k. o. a. 1956. Die Serpuliden-Fauna des siidwestdeutschen Jura. Palaeontographica, Abteilung A, 107,
211-240.

PHILLIPS, J. 1829. Illustrations of the geology of Yorkshire ; or a description of the strata and organic remains of
the Yorkshire coast. Thomas Wilson and Sons, York, 192 pp.
pisera, a. 1987. Boring and nestling organisms from Upper Jurassic coral colonies from northern Poland.
Acta Palaeontologica Polonica, 32, 83-104.

pohowsky, r. a. 1974. Notes on the study and nomenclature of boring Bryozoa. Journal of Paleontology, 48.
557-564.

1978. The boring ctenostomate Bryozoa: taxonomy and paleobiology based on cavities in calcareous
substrata. Bulletins of American Paleontology, 73 (301), 1-192.
portlock, J. E. 1843. Report on the geology of the county of Londonderry, and of parts of Tyrone and Fermanagh.

A. Milliken, Dublin, 748 pp.
quenstedt, F. a. 1858. Der Jura. Tubingen, 842 pp.
rafinesque, c. s. 1815. Analyse de la nature. Palermo, 225 pp.
reineck, h.-e. 1980. Steinkerne in der Entstehung. Natur und Museum , 110, 44-47.

reitner, j. and keupp, H. 1991. The fossil record of the haplosclerid excavating sponge Aka de Laubenfels.

102- 120. In reitner, j. and keupp, h. (eds). Fossil and Recent sponges. Springer, Berlin, 595 pp.
roeding, P. F. 1798. Museum Boltenianum sive catalogus cimeliorum e tribus regionibus naturae quae olim
collegerat Joa. Fried. Bolten M.D. p.d. Pars secunda. Hamburg, viii + 119 pp.
rutzler, k. 1975. The role of burrowing sponges in bioerosion. Oecologia, 19, 203-216.

1978. Sponges in coral reefs. 299-313. In stoddart, d. r. and Johannes, r. e. (eds). Coral reefs: research
methods. Unesco, Paris, 436 pp.

and macintyre, i. G. 1978. Siliceous sponge spicules in coral reef sediments. Marine Biology, 49, 147-159.

FURS1CH ET AL.: JURASSIC REEFS

171

saint-seine, R. de 1951. Un cirripede acrothoracique du Cretace: Rogerella lecontrei n. g., n. sp. Comptes
Rendus Hebdomadaires des Seances de k Academic des Sciences , Paris, 233, 1051-1053.

— 1954. Existence de cirripedes acrothoraciques des le Lias: Zapfella pattei nov. gen., nov. sp. Bulletin de
la Societe Geologique de France, (6), 4, 441-451.

silen, l. 1946. On two new groups of Bryozoa living in shells of molluscs. Arkiv Zoologi, 38B, 1-7.
stewart, r. b. 1930. Gabb’s California Cretaceous and Tertiary type lamellibranchs. Academy of Natural
Sciences of Philadelphia, Special Publication, 3, 1-314.

sun, s. Q. and wright, v. p. 1989. Peloidal fabrics in Upper Jurassic reefal limestones. Weald Basin, Southern
England. Sedimentary Geology, 65, 165-181

taylor, p. d. 1990. Preservation of soft-bodied and other organisms by bioimmuration - a review.
Palaeontology, 33, 1-18.

townson, w. g. 1975. Lithostratigraphy and deposition of the type Portlandian. Journal of the Geological
Society , London , 131, 619-638.

tryon, g. w. 1862. Monograph of the family Teredidae. Proceedings of the Academy of Natural Sciences of
Philadelphia, 13, 453-482.

voigt, e. 1971. Fremdskulpturen an Steinkernen von Polychaeten-Bohrgangen aus der Maastrichter
Tuffkreide. Palaontologische Zeitschrift, 45, 144-153.

— 1975. Tunnelbaue rezenter und fossiler Phoronidea. Palaontologische Zeitschrift , 49, 135-167.

Walter, L. M. 1985. Relative reactivity of skeletal carbonates during dissolution: Implications for diagenesis.

3-16. In schneidermann, n. and Harris, p. m. (eds). Carbonate cements. Society of Economic Paleontologists
and Mineralogists, Special Publication , 36, 379 pp.

warme, j. e. and mchuron, e. j. 1978. Marine borers: trace fossils and their geological significance. 67-118.
In basan, p. b. (ed.). Trace fossil concepts. Society of Economic Paleontologists and Mineralogists, Short
Course , 5, 181 pp.

watson, H. 1930. On the anatomy and affinities of Plicatula. Proceedings of the Malacological Society, London ,
19, 25-30.

werner, w. 1986. Palokologische und biofazielle Analyse des Kimmeridge (Oberjura) von Consolaqao,
Mittelportugal. Zitteliana, 13, 1-109.

Wimbledon, w. a. and cope, j. c. w. 1978. The ammonite faunas of the English Portland Beds and the zones
of the Portland stages. Journal of the Geological Society, London, 135, 183-190.
woodward, H. b. 1895. The Jurassic rocks of Britain. Volume 5. Middle and Upper Oolitic rocks of England
(Yorkshire excepted). Memoirs of the Geological Survey of the United Kingdom, 1-499.
wright, v. p. 1985. Seasonal banding in the alga Solenopora jurassica from the Middle Jurassic of
Gloucestershire, England. Journal of Paleontology, 59, 721-732.

FRANZ T. FURSICH
Institut fur Palaontologie
Universitat Wurzburg
Pleicherwall 1

D-97070 Wurzburg, Germany

TIMOTHY J. PALMER
KAY L. GOODYEAR

Typescript received 5 January 1993
Revised typescript received 30 April 1993

Institute of Earth Studies
University of Wales, Aberystwyth
Aberystwyth, Dyfed SY23 3DB, UK

A NEW CONIFEROUS MALE CONE FROM THE
ENGLISH WEALDEN AND A DISCUSSION OF
POLLINATION IN THE CHEIROLEPIDIAECAE

by K. L. ALVIN, J. WATSON and R. A. SPICER

Abstract. The cheirolepidiaceous pollen-bearing cone Classostrobus comptonensis is redescribed on the basis
of new material from the type locality. A different pollen-bearing cone from the same deposit is described as
a new species of Masculostrobus. The numerous specimens display a range of variation which probably
represents various degrees of maturity. The structure of this new species, which is typically coniferous, is closer
to that found in Taxodiaceae or Araucariaceae than any other extant family. It cannot yet be attributed to any
leafy shoot.

The genus Classostrobus Alvin et al. (1978) was erected for Classopollis- containing cones which
could not be attributed with certainty to any species of leafy shoot. The type species, C. comptonensis
Alvin et al. (1978) was based on specimens from a plant debris bed of Barremian age exposed at the
cliff foot and in the foreshore between Compton Grange Chine (sometimes called Shippard’s Chine)
and Hanover Point (sometimes called Brook Point) in the Isle of Wight (OS Grid Ref. SZ 377840).
The bed contains abundant shoot fragments and wood of the cheirolepidiaceous conifer,
Pseudofrenelopsis parceramosa (Fontaine) Watson (Watson 1977; Alvin et al. 1981; Alvin 1983).
Classostrobus comptonensis was attributed to this plant on the basis of association, and on cuticle
similarities.

Further collecting from the same bed has yielded quite numerous pollen-bearing cones. Among
these were the two new specimens of C. comptonensis from which Taylor and Alvin (1984) obtained
pollen for their study of wall development in Classopollis. The same study, however, indicated that
a different type of male cone was also present, evidenced by its different pollen; it was attributable
to the pollen genus Araucariacites Cookson ex Couper (Batten, pers. comm.). Indeed, it has now
become clear that the great majority of the cones retrieved from the bed belong to this other species.
Of some thirty cones which have been recovered, only three complete, and a few fragments can be
identified as C. comptonensis. Moreover, examination of the original material on which this species
was erected has shown that it was also mixed. It is therefore necessary to emend C. comptonensis
and describe the other cone as a new species of Masculostrobus Seward, a form-genus of coniferous
pollen cones not attributable to any family. In the course of our study, some new information has
emerged concerning the structure of the microsporophyll of C. comptonensis. In the light of this, we
discuss the possible significance of some of the unusual features of the cheirolepidiaceous male cone.

MATERIALS AND METHODS

Most of the specimens have been obtained by bulk breakdown of dried blocks of matrix. The blocks were
soaked in water for several hours after which the material was sieved, washed and, with the material left in
water, searched under a stereoscopic microscope. Cuticles were prepared by using Schulze’s macerating fluid,
followed by 5 per cent potassium hydroxide and mounted either in glycerine jelly for light microscopy or on
stubs using double-sided adhesive tape for scanning electron microscopy. Unmacerated pieces were mounted for
SEM in the same way. Material softened overnight in 2 per cent potassium hydroxide dissolved in 70 per cent
alcohol was used for dissection. All specimens (numbers prefixed V.) are in the Palaeontology Collections,
Natural History Museum, London.

(Palaeontology, Vol. 37, Part 1, 1994, pp. 173-180, 1 pl.|

© The Palaeontological Association

174

PALAEONTOLOGY, VOLUME 37

SYSTEMATIC PALAEONTOLOGY

Family cheirolepidiaceae Takhtajan, 1963
Form-genus classostrobus Alvin et al., 1978

Type species. Classostrobus comptonensis Alvin et al., 1978

Classostrobus comptonensis Alvin et al.

Text-figure 1a-c

Emended diagnosis. Cone spherical to ellipsoidal, with maximum dimensions 14 mm and 12 mm.
Microsporophylls spirally arranged, imbricated, peltate, consisting of a laminate head with centrally
attached stalk 0-3-0- 7 mm in diameter and 1-2 mm long; head rhomboidal, 2-4 mm long,
l-5-3-0mm wide, with acute apex (56-80°), upper portion (exposed on cone surface) thick with
papillate surface and marginal hairs up to 60 //m long; lower portion (covered by heads of lower
sporophylls in cone) thinner and more delicate, becoming extremely thin towards lower edge,
smooth. Abaxial cuticle of upper portion thick, papillose, with cutinized hypodermis, and stomata
in ill-defined rows or more or less scattered; papillae conical, 10-15 //m long, arising singly from
epidermal cells; epidermal cells isodiametric, 20-30 //m wide, with anticlinal walls 4—8 //m thick;
hypodennal cells isodiametric, 20-30 pm wide or, between stomatal rows, elongated, 10-20 //m
wide, stomata sunken, with pit formed by a ring of 5-7 subsidiary cells, each with a papilla up to
1 5 pm long. Abaxial cuticle of lower portion of sporophyll head thinner, smooth, with stomata
confined to upper region; these stomata less sunken and without papillae. Adaxial cuticle of upper
portion of sporophyll head thin, papillate in central region down to attachment of stalk, smooth
elsewhere; cells isodiametric, not well marked. Sporophyll stalk smooth with elongated epidermal
cells 60-80 //m long, 15-30//m wide. Pollen of Classopollis- type, spheroidal with flattened poles,
diameter ( 30— )36( 40) pm x (23— )25(— 27) pm [fifteen grains measured], divided into two unequal
caps by a subequatorial furrow situated at the distal edge of an equatorial belt (8—) 1 1 (—12) pm wide;
9-12 striations in equatorial belt. Pseudopore at distal end 6-11 pm in diameter. Trilete mark at
proximal pole with laesurae 5-7 pm long. Exine thickness 1-3 pn i; equatorial thickening 2-5 pm.
External sculpture fine, of small blunt spines, about 25 elements per //m2. Internal sculpture
vermiculate (see Text-fig. 1a-c).

Types. The holotype is specimen V.591 15 (Alvin et al. 1978, pi. 97, figs 1-7; pi. 98, figs 1-6; text-fig. 1a); from
the Brook Formation (Barremian) between Compton Grange Chine and Hanover Point (OS Grid Ref.
SZ 377840), Isle of Wight, southern England.

Description and discussion. The new material on which our revision of this species is based consists
of two more or less complete cones (V. 63682 and V. 63683) and one broken specimen (V. 63684). The
better preserved cone (V. 63682) has allowed a more detailed description of the microsporophyll
surfaces and cuticles.

The number of pollen sacs on the microsporophyll, given as three in the original description, was
based unfortunately on the small cone illustrated by Alvin et al. (1978, pi. 96, figs 8—9 ; text-fig.
1b-d), a specimen which belongs to the new species described below. The number of pollen sacs in
Classostrobus comptonensis is therefore unknown. However, on the basis of some evidence from one
of the new cones (V. 63682), part of which we dissected after softening in alcoholic potash, as well
as from some microsporophylls removed from the broken specimen (V. 63684), we believe that the
number was probably more than two. The dissected cone indicated a very large number of yellowish
bodies, presumably remains of pollen sacs, close to the axis, but it was not possible to determine
their individuality or how they were associated with the individual microsporphylls. Some
microsporophylls removed from the broken cone showed patches of pollen, or small, very indistinct
depressions on the lower part of the adaxial side, which suggested that the number might have been

ALVIN ET A L. : WEALDEN CONIFEROUS CONE

175

text-fig. I . Conifer pollen cones from between Compton Grange Chine and Hanover Point, Isle of Wight,
Brook Formation (Barremian). a-c, Classostrobus comptonensis Alvin et al. ; V. 63684; isolated micro-
sporophylls showing shallow depressions (indicated by dotted lines) below the stalk, possibly representing
attachments of pollen-sacs; lower edge (shown as a broken line) is probably damaged, d-j, Masculostrobus
vectensis sp. nov. d-i V. 63687 to V. 63692 respectively; a selection of cones showing the range of size and form;
in each, the two sides of the specimen are shown; pollen-sacs are stippled. J, V. 63690, typical grouping of
pollen-sacs obtained by softening and squashing the cone shown in Text-figure 1g.

176

PALAEONTOLOGY, VOLUME 37

three (Text-fig. 1a-c). We envisage that the pollen-sacs were attached to the lower part of the
sporophyll and were probably elongated parallel to the stalk. This agrees essentially with Kirchner’s
(1992) interpretation of the structure of the microsporophyll in Hirmeriella escheri (Heer) Kirchner,
except that the number of pollen-sacs is about ten per sporophyll in that species.

Family unknown

Form-genus masculostrobus Seward, 1911
Type species. Masculostrobus zeilleri Seward, 1911.

Masculostrobus vectensis sp. nov.

PI. 1, figures 1-7; Text-figure 1d-j

Diagnosis. Cone shortly ellipsoidal to almost spherical (2-2-)3-4(^P6) mm long, ( l -3— )2-7(— 3-7) mm
wide. Microsporophylls spirally arranged, imbricate, peltate, each consisting of a laminate head and
a sub-centrally placed stalk; exposed part of head typically pointed-elongate, c. 1-5 mm long,
08 mm wide but varying greatly in size more or less proportionally to the size of cone; apex acute
(40°-)54°(-75°) ; scarious border consisting of cells elongated almost at right angles to the edge and
becoming a single layer thick with the marginal cells drawn out into free or contiguous hairs. Cuticle
of abaxial side thin, showing slightly elongated cell outlines; stomata absent. Pollen sacs borne on
lower half of microsporophyll, typically three in number, c. 04 by 0T5 mm, cylindrical or
tapering slightly towards the axis. Pollen of Araucariacites type, spherical, inapperturate,
(48-0-)57-9(-64-8) pm in diameter [20 grains measured]; exine finely granular with elements
c. 0-4 jum in diameter and minutely rough surface.

Types. The holotype is specimen V. 591 17 (Alvin et al. 1978, pi. 96, figs 8-9; text-fig. 1c-d); from the Brook
Formation (Barremian), between Compton Grange Chine and Hanover Point (O.S. Grid Ref. SZ 377840), Isle
of Wight, southern England.

Description and discussion. The quite numerous specimens of this new species present a considerable
range, not only in cone and sporophyll size, but also in the plane of compression (PI. 1, fig. 1 ; Text-
fig. 1d-i). This last variation no doubt relates to the originally almost spherical form. Although the
great majority of specimens appear to be fully mature, a few are considerably smaller than average
in size and have smaller, more tightly adpressed sporophylls in which the scarious margin is less
conspicuous (e.g. Text-fig. 1g, i). Pollen has been obtained in greatly varying quantity from several
specimens. In some cones, particularly the smaller ones with closely adpressed microsporophylls,
there was a strong tendency for the pollen to adhere together in masses apparently representing
portions of pollen-sacs. Sometimes these pollen masses are accompanied by remains of a bounding
membrane and usually abundant Ubisch bodies (orbicules). We interpret these specimens as being
immature and envisage that maturation was accompanied by expansion of individual micro-
sporophylls. This conforms with what happens typically in most living conifers.

EXPLANATION OF PLATE 1

Figs 1-7. Masutostrobus vectensis sp. nov. Between Compton Grange Chine and Hanover Point, Isle of Wight;
Brook Formation, (Barremian). 1, V. 63685; cone; x 18. 2-4, V. 63686, tip of microsporophyll showing
extreme apex and fimbriate margin; x 100. 3, detail of marginal hairs; x 650. 4, SEM view of head of
microsporophyll with damaged margin and tip, showing epidermal cells; x 120. 5-7, V. 63687; views of
pollen grain from cone shown in Text-figure 1e. 5, SEM view; x 1250. 6, SEM view of surface; x 10000. 7,
light microscope view; x750.

PLATE 1

ALVIN et al., Masculostrobus

178

PALAEONTOLOGY, VOLUME 37

The specimen shown in Text-figure 1g was judged to represent a young cone. After being softened
in alcoholic potash, portions were squashed on a microscope slide and yielded numerous whole
pollen sacs; these showed a strongly grouped arrangement (most commonly in threes). This
probably reflects the number of pollen sacs on individual microsporophylls (Text-fig. 1j) and thus
agrees with the observations originally made on the cone that we now designate as the type specimen
(Alvin et al. 1978, text-fig. lc-D).

The sporophyll margin is a distinctive feature, though this is much less conspicuous in the small,
probably unexpanded cones, and also in the small sporophylls close to the apex and the base of
larger specimens. In unmacerated sporophylls the marginal band becomes translucent as it thins to
a single layer of elongated cells at the edge, which are then extended into fringe-like hairs (PI. 1, figs
2-3). In many examples the sporophyll edge appears smooth and slightly wavy, but this is believed
to be due to the breaking off of the hairs; similarly, the extreme tip is usually broken off (PI. 1, fig.

4).

Cuticles are thin and difficult to prepare, but the abaxial surface of the cleaned but unmacerated
sporophyll viewed directly by SEM sometimes shows slightly elongated, bulging epidermal cells
(PI. 1, fig. 4).

Typical pollen grains are shown in Plate 1, figures 5 and 7, and the exine sculpture in Plate 1,
figure 6. The pollen characters given in the diagnosis are based mainly on samples obtained from
the type specimen and two others (V. 63688 and V. 63689; Text-fig. 1e-f). The grains from the former
had a slightly larger average size than those from the other, though the size ranges overlapped and
the maximum size was the same in both. The cone with the slightly larger average size also had a
higher proportion of free grains, probably suggesting it was more mature. The degree of roughness
of the exine granules is somewhat variable. Although there is no clear evidence of an aperture,
squashed or otherwise damaged grains sometimes indicate that there may have been a slightly
thinner area of wall.

The structure of Masculostrobus vectensis does not itself give any clear indication of family
affinity; hence we classify it in this form-genus which was erected by Seward (1911) for male cones
of unknown affinity. Cones of the same basic structure are found in a number of extant conifer
families. However, the spiral arrangement of the sporophylls, the presence of more than two pollen
sacs on the sporophyll and non-saccate pollen represent a combination of features which seems to
us more reminiscent of Taxodiaceae or Araucariaceae than of other families. However, the delicate
cone structure makes it dissimilar to most living Araucariaceae.

Although remains of plants other than Pseudofrenelopsis parceramosa are rare in the deposit and
represented only by fragments, leaves and leafy shoots of some three or possibly four other confers
are present. These are under investigation.

DISCUSSION

Pseudofrenelopsis parceramosa , represented by ultimate shoots, woody twigs and wood, is often
very well preserved and abundant in the same bed which has yielded both of the cones described
above (Alvin et al. 1981; Alvin 1983). Other plants, in contrast, are infrequent, fragmentary and
often not well preserved. It is surprising therefore that specimens of Classostrobus comptonensis ,
attributable on good evidence to Pseudofrenelopsis parceramosa , are found only rarely, compared
with Masculostrobus vectensis. It is even more surprising in view of the comparatively robust
structure of the former and the delicate nature of the latter. We tentatively suggest that this may
relate to differences in the biology of the two species with regard to pollen shedding and perhaps
pollination.

The parent tree of Masculostrobus vectensisis is likely to have produced pollen in numerous,
short-lived cones, which expanded late in their development and were then shed. Such a scenario
is usual in living conifers and indeed in many other wind-pollinated plants. Classostrobus
comptonensis , like most of the fifteen or so cheirolepidiaceous male cones which have been described
(Watson 1988), is unusual in being comparatively robust in structure with the microsporophyll as

ALVIN ET AL.: WEALDEN CONIFEROUS CONE

179

thickly cutinized on the exposed abaxial side as the leaf or photosynthetic stem of the same plant.
Thick cutinization of the microsporophyll is unusual amongst living conifers. Araucaria araucaria ,
which has large, fairly long-lasting male cones, has strong cutinization of the microsporophyll, but
the thicker abaxial cuticle is only about half as thick as the leaf cuticle. Whilst the unusual thickness
of the microsporophyll cuticle in a cheirolepidiaceous species may be merely part of the generally
xeroinorphic character of the whole plant, it suggests that the male cone may have been long-lasting
rather than ephemeral and, if so, may have shed pollen slowly or perhaps only at intervals over a
long period. The fact that all but two or three of the known cheirolepidiaceous male cones have been
found in attachment to shoots seems to support this hypothesis. Amongst other fossil conifers
attachment is encountered less frequently, presumably because, as in living conifers, the short-lived
cones were shed rapidly after pollen dispersal. Is it possible, therefore, that there may have been a
mechanism for slow or controlled pollen release in Cheirolepidiaceae? The cone structure gives little
clue, but it is noteworthy that in Classostrobus comptonensis considerable quantities of pollen are
found adhering to the overlapping surfaces of the microsporophylls. Such pollen could conceivably
have been released slowly from between these hairy surfaces by small movements of the
microsporophylls, made perhaps in response to small changes in atmospheric humidity in much the
same way as in the seed-cone scales in living Pinaceae. Many examples of controlled release of
disseminules are known among living plants such as the xerochasic release of spores from many
moss capsules (Ingold 1965) and the hydrochasic mechanism releasing seeds in some desert plants
(Van der Pijl 1969).

Recently, on the basis of their new study of Hirmeriella muensteri , Clement-Westerhof and Van
Konijnenburg-Van Cittert (1991) have suggested that the pollen might have alighted and
germinated on the female cone-scale and that the pollen tubes then grew towards the ovules. This
behaviour is well known in a number of present-day conifers (Doyle 1945) but is not correlated with
any obvious modification of the male cone, although it is associated in Pinaceae (Tsuga) and
Podocarpaceae ( Saxegolhaea ) with loss of pollen sacci.

The extraordinary elaborate pollen wall structure in Classopollis and other circumpolles, recently
reviewed by Pocock et al. (1990), has led these authors to suggest that the group might have
advanced towards entomophily in the Jurassic and Cretaceous. Nevertheless they see some of the
wall features as being more especially concerned with permitting volume changes in the grain during
transit through the arid environment. If, as we suggest, the pollen remained viable in the mature
male cone for a long period after its initial release from the pollen sacs until, perhaps only under
certain conditions, being transported by the pollination vector, resistance to desiccation and
longevity could have been features of special importance.

Acknowledgement . We are grateful to Mr Sinclair Stammers of the Department of Biology, Imperial College,
London, for photographing the specimen in Plate 1, figure 1.

REFERENCES

alvin, K. l. 1983. Reconstruction of a Lower Cretaceous conifer. Botanical Journal of the Linnean Society , 86,
169-176.

-Fraser, c. J. and spicer, r. a. 1981. Anatomy and palaeoecology of Pseudofrenelopsis and associated
conifers in the English Wealden. Palaeontology, 24, 759-778.

— spicer, r. a. and watson, J. 1978. A Classopollis- containing male cone associated with Pseudofrenelopsis.
Palaeontology , 21, 847-856.

clement-westerhof, j. a. and van Konijnenburg-Van cittert, j. h. a. 1991 Hirmeriella muensteri : new data
on the fertile organs leading to a revised concept of the Cheirolepidiaceae. Review of Palaeobotany and
Palynology , 68, 147-179.

doyle, J. 1945. Developmental lines in pollination mechanisms in the Coniferales. Scientific Proceedings of the
Royal Dublin Society, 24, 43-62.

ingold, c. T. 1965. Spore liberation. Oxford University Press, Oxford, 210 pp.

180

PALAEONTOLOGY, VOLUME 37

kirchner, M. 1992. Untersuchungen an einigen Gymnospermen der frankischen Rhat-Lias-Grenzschichten.
Palaeontographica , Abteilung B. 224, 17-61,

pocock, s. a. j„ vasanthy, G. and venkatachala, b. s. 1990. Pollen of circumpolles - an enigma or
morphotrends showing evolutionary adaptation. Review of Palaeobotany and Palynology , 65, 179-193.
seward, a. c. 1911 The Jurassic flora of Sutherland. Transactions of the Royal Society of Edinburgh , 47,
643-709.

takhtajan, a. L. 1963. [Gymnospenns and angiosperms]. Osnovy Paleontologii, 15, 1-1743. [In Russian].
taylor, r. N. and alvin, k. l. 1984. Ultrastructure and development of Mesozoic pollen: Classopollis.
American Journal of Botany , 71, 575-587.

van der pul, L. 1969. Principles of dispersal in higher plants. Springer-Verlag, Berlin, 154 pp.
watson. j. 1977. Some Lower Cretaceous conifers of the Cheirolepidiaceae from the U.S.A. and England.
Palaeontology , 20, 715-749.

1988. The Cheirolepidiaceae. 382-749. In beck, c.b. (ed. ). The origin and evolution of gymnospenns.
Columbia University Press, New York, 504 pp.

K. L. alvin
R. A. SPICER

Department of Earth Sciences
University of Oxford
Parks Road
Oxford OX1 3PR, UK

J. WATSON

Departments of Environmental Biology and Geology
Typescript received 6 October 1992 University of Manchester

Revised typescript received 9 June 1993 Manchester M13 9PL, UK

EIDER, SHELDUCK, AND OTHER PREDATORS, THE
MAIN PRODUCERS OF SHELL FRAGMENTS IN THE
WADDEN SEA: PALAEOECOLOG IC AL
IMPLICATIONS

by GERHARD C. CADEE

Abstract. Seventy five per cent by weight of the > 2 mm carbonate fraction of Wadden Sea sediments consists
of fragmented shells, thirty per cent > 8 mm and forty five per cent in the 2-8 mm fraction. Eiderducks
(Somateria mollissima) feed mainly on mussels ( Mytilus edulis) and cockles (Cerastoderma edule). Shells are
crushed internally to fragments with a size-range from < 01 to 8 mm, twenty per cent were < 1 mm, sixty per
cent 2-8 mm. One-third to one-half of the fragments in the 2-8 mm fraction in the sediments are due to eider
predation alone. Other birds, crabs and fish probably produce the remaining fragments of this fraction.
Shelduck ( Tadorna tadorna) feed on the small gastropod Hydrobia ulvae; a varying amount (seventeen to thirty
two per cent by weight) of shells was found intact in their faeces, but the remainder is fragmented. Around forty
per cent by weight of Hydrobia shells in the Wadden Sea sediments is broken. This can be attributed to shelduck
and other predators (e.g. knot) feeding on Hydrobia. Fragments in the > 8 mm fraction may also be produced
by predators (shore crabs, oystercatchers). Physical destruction plays a minor role in the Wadden Sea. Shell
fragmentation cannot be used as a measure of water turbulence. The high percentage of shell fragments
indicates high predation pressure. However, the use of shell fragmentation to estimate predation pressure in
fossil faunas is not possible, because some predators leave one (oystercatchers) or both valves ( Asterias ) intact.
Despite high fragmentation fidelity of the death assemblage to the living fauna of the Wadden Sea is high.
Physical destruction would leave only fragments of durable skeletons with low fidelity to the living fauna.

Fragmentation of shells in marine sediments may be due to biological or physical processes.
Early workers for example Woodward (1875), Verrill (1882) and Walther (1893, 1910), stressed the
importance of shell crushing by predators, mainly crustaceans and fish. Later, more attention was
paid to mechanical fragmentation, which was also studied experimentally (Klahn 1932; Chave
1960; Force 1969). Van Straaten (1952, 1956) introduced a ‘crush factor’ (the percentage of broken
shells of the entire shell material of one species) to be used as an indicator of the degree of wear of
the shell material during transport, a method also advocated by Ager (1964, p. 199). A comparable
correlation between fragmented shells and water turbulence has been suggested by Vokes (1948),
Bissell and Chilingar (1967), and Link (1967). Studies of experimental abrasion and fragmentation
of skeletal elements are still popular, for example, Kidwell and Baumiller ( 1 990), Kotler et al. ( 1 992).

Biological factors in shell fragmentation have been stressed by Ginsburg (1957), Schafer (1962)
and Trewin and Welsh (1976). Carter (1968) and Vermeij (1978, 1987) summarized an extensive
literature dealing with shell fracturing by molluscan predators. The dual nature of shell
fragmentation is now well accepted (Chave 1964; Pilkey 1964, 1969; Swinchatt 1965; Cadee 1968;
Palmer 1977; Dodd and Stanton 1981), but it makes interpretation of fragmented shells difficult.
Although Dodd and Stanton (1981) suggested that the products from these two processes are
recognizable, Powell et al. (1989) more realistically remarked that the interpretation of the
fragmented portion of a molluscan death assemblage is often difficult. Parsons and Brett’s (1991)
statement that ‘fragmentation is clearly a measure of environmental energy, with the exception of
biologically mediated breakage’ gives little hope for solving the problem. Their suggestion that

IPalaeontology, Vol. 37, Part 1, 1994, pp. 181-202, I pl.|

© The Palaeontological Association

182

PALAEONTOLOGY, VOLUME 37

environmental energy is the main cause of shell fracturing is counter to that of Cadee (1968) and
Dodd and Stanton (1981), who suggested biological factors such as predation are most important
in shell breakage. Only in a few cases can shell fracturing be attributed mainly to physical
destruction in the surf zone of a sandy (Hollmann 1968) or a pebbly (Trewin and Welsh 1972) beach.

Shell fragments produced by predators are sometimes very predator specific (Yernreij 1978,
1987), but later abrasion and activity of boring organisms may make these fragments similar to
those produced by physical processes (Schafer 1962). Quantitative separation of biological and
physical processes in fragmentation therefore does not seem possible. Fracturing by water energy
will be largely caused by waves and breakers and is thus confined to the coastal zone, but in this
coastal zone predators (mainly birds) are also active (this paper). Shell fragments may be produced
sublitorally by predators (fish, crustaceans, birds) and transported as fragments to the coast.
Sublittoral currents may not be strong enough to fragment shells (Schafer 1962; Feige and Fiirsich
1991). Flemming et al. (1992) stated that normal tidal currents are not strong enough even to
transport shells in the Wadden Sea: subtidal shell beds are formed there by a largely passive
gravitational transport in the wake of channel migration.

With the increasing quantitative knowledge of food webs in the recent marine environment, it
now becomes possible to estimate the amount of the total biomass of molluscs broken by different
predators for some areas. This can be compared with the actual ‘crush factor' found for the empty
shells in the sediment and may enable an estimation of the importance of biological versus physical
shell fragmentation. Eisma and Hooft (1967) were the first to apply this method. They compared
the amount of shell fragments present in Recent Dutch open-sea coastal North Sea sediments and
concluded that 50 to 100 per cent of the fragments were due to flatfish (plaice) predation. For the
Holocene Dutch open-sea, coastal sediments they estimated that at least 10 per cent, but probably
much more of the molluscan fragments were due to plaice predation. An admixture of fragments
from Eemian sediments, breakage by other predators, and some physical fragmentation in the surf
zone could account for additional fragmentation. Similar quantitative work has not been carried
out since.

In this paper I concentrate on the Dutch Wadden Sea. In the last three decades a large amount
of data has been collected on macrobenthos biomass and production in this area (Beukema 1976,
19826, 1989; Dekker 1989; Dankers et al. 1989) and on predators, mainly birds, compiled by
Hulscher (1975), Swennen (1975) and Smit (1981).

Data are presented on shell fracturing in the Dutch Wadden Sea by two duck species, shelduck
( Tadorna tadorna ) and eider ( Somateria mollissima). Both species ingest whole molluscs and crush
the shells in their gizzard, so that they can digest the meat. Shelduck feed on small molluscs, such as
Hydrobia idvae, and small worms and crustaceans (compilation in Bauer and Glutz 1968), ‘sieving’
those > 2 mm from the sediment with their beak (Thompson 1982). They use different methods,
depending on water depth, to collect food (Bryant and Leng, 1975, fig. 4), and leave characteristic
feeding traces on the sediment (Cadee 1990), but do not dive. Eider feed mainly on larger molluscs
like the cockle ( Cerastoderma edule ) and mussel ( Mytilus edulis ) (see Swennen 1967), which are
collected by diving. These data are compared with available data on production of these molluscs
(Dekker 1979; Beukema 1980, 1982u; Dankers et al. 1989) to estimate the fraction broken. This is
compared with the actual percentage of broken shells of the same species in Wadden Sea sediments.
From this comparison it is possible to estimate the role of these ducks in the production of shell
fragments in Wadden Sea sediments.

MATERIAL AND METHODS

Faeces of eider were collected from several locations on the dike along the southern part of Texel
(Text-fig. 1), where a small part of the eider population roosts (most eider ‘roost’ on open water,
Swennen 1976, 1991). Such faeces usually consist of the undigestible parts of one food item only,
such as mussels and cockles, and when these are scarce, shore crabs ( Carcinus maenas ), or
periwinkles ( Littorina littorea) (Cadee 1991). To get rid of the organic coating of the faeces they

CADEE: RECENT SHELL FRAGMENT PRODUCERS

183

text-fig. 1. Location of sampling stations (mostly in 1992), numbering according to depth (from station 1,

High Water line, to 20 m depth for station 9).

were shaken with water in glass beakers, and the water decanted. This was repeated several times
until the decanted water became more or less clear. Part of the smaller particles in the faeces,
roughly < 50 //m, were inevitably also removed during cleaning. The size distribution of the shell

184

PALAEONTOLOGY, VOLUME 37

fragments was measured by dry sieving over a series of sieves with mesh sizes ranging from 0-1 to
8 mm.

Faeces of shelduck were collected on the tidal flat in the Mokbay (Text-fig. 1) where shelduck
were observed feeding. Their faeces are easy to recognize by their size, and they are often full of
hydrobiids. These faeces were treated in the same way as the eider faeces, but because shelduck
faeces also contain varying amounts of larger sand and shell grains picked up during feeding, the
size frequency distribution of Hydrobia fragments could not be measured by sieving. Some faecal
counts were made of the number of broken (only top fragments were counted) and entire Hydrobia
shells. The crush factor (weight of all Hydrobia fragments as a percentage of total weight of
Hydrobia shells) was also determined.

Bottom samples were collected at a number of locations near Texel (Text-fig. 1). A Van Veen grab
was used for sublittoral sites and intertidal sites were collected by hand. They were sieved over 1,
2 and 8 mm sieves and a ‘crush factor’, i.e. the weight percentage of broken shells of the total
amount (Van Straaten 1956), was measured from 1000-2000 particles in the > 2 mm fraction. Also
crush factors for the dominant species such as cockle and mussel were measured separately. For the
small Hydrobia the fraction 1-2 mm also was used to estimate this crush factor.

Eider faeces

Fortunately an individual eider usually feeds on one food item at a time. Its faeces therefore contain
only material of one food species. In the faeces studied, all mussels and cockles were crushed. In
faeces of eider feeding only on Littorina , a few intact shells were observed. The shell fragments
produced were angular and sharp edged (PI. 1). The size-frequency distributions of fragments in
faeces from the Wadden Sea were similar for the three different types of food: Mytilus,
Cerastoderma and Littorina (Text-fig. 2). The bulk of the fragments was in the fraction 2-4 mm;
very few fragments were larger than 8 mm, approximately 20 per cent was smaller than 1 mm, and
60 per cent was in the 2-8 mm range. Eider faeces collected in the Baltic at Aland, where they were
feeding on smaller (up to 30 mm in length) and thin-shelled Mytilus , had a different size frequency
distribution (Text-fig. 3). Fragments had a peak in the 1-2 mm fraction, 60 per cent were 1 mm, and
only 6 per cent > 2 mm.

Shelduck faeces

Shelduck faeces consisted mainly of Hydrobia shells. Remains of additional prey items were of small
bivalves, Macoma balthica , Cerastoderma edule and Abra tenuis , and varying amounts of worms
and crustaceans. The remainder consisted of coarse sand grains and a few fragments of large
bivalves picked up by shelducks during feeding. These grains might help in grinding Hydrobia shells.
However, not all Hydrobia shells appeared to be broken. In fact, a whole range from small
fragments to entire shells could be found (Text-fig. 4). A number of Hydrobia were even still alive
after passage through the shelduck, which takes one to two hours. This was not reported earlier and
is a possible factor in short-distance dispersal of this gastropod (Cadee 1989). A comparable
survival of 10 per cent of gastropods consumed in the intertidal of the Pacific by the fish Asemichthys
taylori was reported by Norton (1988). Apparently an intact shell plus its operculum are a sufficient
barrier against digestion. Droppings (all the faeces produced at one moment) contain remains of a
few hundred to over a thousand Hydrobia shells (Table 1). Crush factors for Hydrobia shells in the

EXPLANATION OF PLATE 1

Different fractions of shell fragments from eider faeces collected on Texel. Rows from top to bottom are
4-8 mm, 1 2 mm, 0-5-0-7 mm and 01-0-2 mm fractions, respectively. Left column (figs 14): Mytilus edulis ,
25 January 1988. Right column (figs 5-8): Cerastoderma edule , 14 May 1991. All figures x 2.

PLATE 1

CADEE, shell fragments

186

PALAEONTOLOGY, VOLUME 37

grainsize (micron)

text-fig. 2. Size frequency distribution of shell fragments in eider faeces collected on the dike of Texel.
Numbers in legend indicate date (month/day) in 1991 (except 1988 for 25/1). Faeces comprised Mytilus (9/5
and 25/1), Cerastoderma (11/5 and 14/5), and Littorina (9/4 and 12/5) fragments.

faeces vary from 68 to 83 per cent (Table 1). Faeces in which high amounts of soft food items were
observed, mainly polychaetes, had lowest crush factors.

WADDEN SEA BOTTOM SAMPLES

In most cases all the sand grains passed the 1 mm sieve leaving only shells and a variable amount
of peat fragments on the sieve. The shell carbonate fraction > 2 mm accounted on average for
95 per cent (range 90-1-99-3 per cent) of bivalves. Gastropods (mainly Littorina littorea) formed
only 3 per cent (range 0-3-4-6 per cent), the rest - 2 per cent (0- 1-5-3 per cent) - consisted largely of
barnacle fragments (Text-fig. 5). On average 2-7 per cent of the fragments, all belonging to bivalves,
could not be identified to the species level. Species composition (Text-fig. 6) indicates that
Cerastoderma edule and Mytilus edulis are the main component (respectively 44-7 and 35-7 per cent)
of the fraction > 2 mm. The small Hydrobia was a minor component of the > 2 mm fraction.

The deepest stations (stations 7-9, Text-fig. 1), near the inlet, contained on average 1 9-6 per cent
Spisula sp., mainly S. subtruncata , a North Sea species occurring in dense patches offshore (Eisma
1966; Offringa 1991). Only occasionally this species settles in the Wadden Sea and may survive
during mild winters, but dies in periods of ice formation on the tidal flats (Cadee and Hegeman
1975). High percentages of Spisula shells in Wadden Sea sediments indicate North Sea influences
(as mentioned already by Krause 1950). Stations 7 and 9 (Text-fig. 1), moreover, contained the
gastropod Bittium reticulation , which is a fossil of Eemian age in The Netherlands (Van Regteren
Altena 1937; Van Regteren Altena et al. 1955). This indicates that Eemian deposits are eroded in
the deep inlet area. Indeed Eemian deposits were found exposed in the deepest inlet channels below
c. 20 m depth by Sha (1989). Therefore some of the other shells, including Cerastoderma and
Mytilus , which were also abundant in the Eemian (Spaink 1965), may be of Eemian age. For a

CADEE: RECENT SHELL FRAGMENT PRODUCERS

187

gramsize (micron)

text-fig. 3. Size frequency distribution of Mytilus fragments in eider faeces from Aland, Baltic Sea, in 1991.

comparison of shell fragments in bottom deposits and the feeding activity of eider and shelduck in
the Wadden Sea, I will only consider to stations 1-6 (Text-fig. 1 ) in the Wadden Sea proper, where
there is less admixture of North Sea and fossil shells.

Crush factors in all bottom samples were high: around 70 per cent for all species together (Text-
fig. 5). Crush factors were highest in the intertidal and shallow Wadden Sea stations (1-3, Text-fig. 1 )
decreasing with depth in the gully stations (4-6, Text-fig. 1), and still lower in the stations in the
deeper inlet and just outside (stations 7-9). Near the High Water line (station 1 ) sorting occurs due
to wave action (Van Straaten 1954, p. 21). The coarse samples had a higher content of entire shells
than the samples with mainly finer material. Four separate samples therefore were taken here to
cover this range and the results were pooled : the average crush factor for this pooled sample (station
1) is comparable to that of the other shallow stations (stations 2-3).

Crush factors varied between species (Text-fig. 7). They were high for typical Wadden Sea
molluscs such as Mytilus , Cerastoderma and Littorina , and lower for the North Sea bivalve Spisula
sp. Typical Wadden Sea species also have relatively high crush factors in samples outside the
Wadden Sea.

HYDROBIA PRODUCTION AND ITS CONSUMPTION BY SHELDUCK

The average number of shelduck occurring in the Dutch Wadden Sea is 21 000 (Smit 1981). Annual
consumption by shelduck in the Dutch Wadden Sea is estimated at 7-1 x 108 g ash-free dry weight
(afdw), being 7 per cent of the consumption of all carnivorous birds (Smit 1981). As shelduck feed
only on tidal flats (120000 ha for the Dutch Wadden Sea) and not in deeper water, their annual
consumption per m2 tidal flat was estimated at 0-59 g afdw (Smit 1981). Dekker (1979) measured
in 1978 an annual production for Hydrobia of 2-0 g afdw on the Balgzand, a tidal flat area in the
western Dutch Wadden Sea, and higher values of 35 5 g afdw in the Mokbay. However, for the
whole western Dutch Wadden Sea this value is probably lower ( c . 0-6 g afdw m“2 tidal flat, Beukema
1981). Year to year variation is large, Balgzand biomass in 1981/1983 amounted to 0-8 g afdw m"2
(Dekker 1989) and using a Production/ Biomass (P/B) ratio of 0-7 (Dekker 1979) this gives an
annual production of only 0-56 g afdw m 2 as compared with 2-0 in 1978.

text-fig. 4. Hydrobia ulvae from shelduck faeces ranging from living specimens to small fragments. The largest

shell is 4 mm in length.

station

|| biv. entire
(2) biv. fragm.
g§ gastr. entire
[2] gastr. fragm.
(23 rest

text-fig. 5. Composition of > 2 mm fraction of Wadden Sea bottom samples; stations arranged according to
depth from High Water (1) to 20 m depth (9). Percentages all by weight.

If shelduck were feeding only on Hydrobia , they could consume the average annual intertidal
production of 0 6 g afdw m”2 of this species. They fed for a large part on Hydrobia as could be
judged from the faeces I collected in Mok Bay, but also took additional prey items (this paper;
see also Bauer and Glutz 1968). Assuming that their diet consists for 50 per cent of Hydrobia and
taking into account that on average 75 per cent of the Hydrobia shells are broken inside shelducks
(Table 1) we can estimate that 37-5 per cent of the shells of the intertidal Hydrobia population are
broken by shelduck. This is near the figure for the crush factor for Hydrobia observed in intertidal
Wadden Sea sediment samples (stations 0-3, average 43 per cent; Table 2). Taking into account
that other birds also consume and crush Hydrobia , e.g. knots ( Calidris canutus) (Piersma et al.
1994; Dekinga and Piersma 1993), I arrive at an even higher potential crush factor than was

CADEE: RECENT SHELL FRAGMENT PRODUCERS

189

0$

»

0

C

0

a

L

0

a

100
so
60
40
20
0

1 23456789

station

0 Cerastodenna

□ Mylilus

□ Macoma
0 Spisula

□ rest bivalves
g§ Littorina

text-fig. 6. Species composition of the molluscan fraction (>2mm) of the same bottom samples as

Text-figure 5.

-o- Mytilus

Cerastoderma
■+ Spisula
-o- Littorina

text-fig. 7. Crush factor (weight percentage of fragments) for four species.

observed. This might indicate that the production estimates for Hydrobia are too low. Certainly,
production of Hydrobia shows large variations from place to place and from year to year.
Moreover, Hydrobia migrates (Dekker 1979) making estimation of its production difficult.

Dekker (1989) observed higher Hydrobia numbers and biomass in the subtidal parts of the
western Wadden Sea, on average 4-9 g afdw m-2. Using again his P/B ratio of 0-7 I arrive at a

190

PALAEONTOLOGY, VOLUME 37

table 1. Crush factors for Hydrobia shells in Shelduck faeces from Mokbay expressed as percentage weight
fragments of total weight of Hydrobia shells and as percentage number of top fragments of entire shells plus
top fragments.

Date

Entire

shells

(number)

Entire

shells

(g)

Top

fragm.

(number)

Top

fragm.

(g)

Crush
factor
by weight

Crush
factor
by number

19 May 1988

277

1162

875

2-512

68-4

76-0

8 June 1988

77

0-294

530

0-941

76-2

87-3

30 June 1988

25

0-055

560

0-262

82-8

95-7

7 Nov. 1988

52

0-167

135

0-390

700

72-2

table 2. Crush factor for Hydrobia shells in Wadden Sea sediment samples (weight of shell fragments as

percentage of weight of all shells in fraction l-2mm+>2mm). Station 0: additional samples of drift
consisting only of Hydrobia shells, Wadden Sea coast, Texel.

Station

no.

Total

weight

(g)

Weight of
fragments
(g)

Crush

factor

0

1 666

0-781

46-9

1

1-113

0-486

43-7

2

1-187

0-486

40-9

3

0-270

0T1 1

41-1

4

0-676

0160

23-7

5

0-907

0-162

17-8

6

0-463

0-103

22-3

7

0-822

0-267

32-5

production of 3-4 g afdw m“2 for the subtidal areas which comprise c. 50 per cent of the Wadden
Sea and where shelduck do not feed. Subtidal predation on Hydrobia has not been quantified.
Potential subtidal predators that crush Hydrobia shells are shore crabs, shrimps, flatfish and gobies.
But predation pressure by these predators is apparently less here accounting for a relatively low
crush factor of Hydrobia shells in subtidal sediment samples (stations 4-7, average 24 per cent.
Table 2).

I conclude that no physical factor in shell fracturing is necessary to account for the broken
Hydrobia shells in intertidal Wadden Sea sediments; this probably also holds for subtidal sediments.

COCKLE AND MUSSEL PRODUCTION AND EIDER CONSUMPTION

Swennen (1976) estimated the average number of eider in the Dutch Wadden Sea at 63000.
Numbers vary seasonally and consist of a breeding population of about 6000 pairs plus a higher
number of non-breeding summer visitors, to a maximum of almost 170000 in winter, mainly due
to immigration from the Baltic population. In the Dutch Wadden Sea the food of the eider consists
of 40 per cent mussels, 40 per cent cockles and 20 per cent other prey (Swennen 1976) and annual
consumption for the Dutch Wadden Sea amounts to c.1-2 g afdw m“2 or 3-2 x 10R kg for the entire
Dutch Wadden Sea of 2600 km2 (Swennen 1976, 1991). Annual cockle and mussel consumption by
eider each amounts to 0-4 x 3-2 x 10R kg afdw = 1-28 x 10R kg. For cockles eaten by eider, shell
carbonate weight is c. twenty times the ash-free dry weight of the meat (Swennen 1976); this gives
an annual deposition of cockle carbonate of 25-6 x 10s kg in the form of shell fragments.

CADEE: RECENT SHELL FRAGMENT PRODUCERS

191

Cockle production varies due to large year to year variations in the cockle population; only few
years produce strong enough settlements leading to strong year-classes that in turn form the bulk
of the cockle biomass (Beukema 1976, 1982a, b). Average annual shell carbonate production was
estimated by Beukema (1982) at 156 x 10” kg for the intertidal and c. 10 per cent of this amount for
the subtidal Wadden Sea. Eiderducks fragment therefore on average c. 15 per cent of the cockle
carbonate production.

For mussels, data are only available for the annual average biomass present in the western
Wadden Sea (14-7 x lO*5 kg afdw, Dankers et al. 1989). As no data are available on production, I will
assume a Production/Biomass (P/B) ratio of one, also used by Beukema (1981) for adult mussels.
This value probably holds for subtidal mussels, but P/B is lower in intertidal areas decreasing from
one in juveniles to one-tenth in ten-year-old mussels (Thompson 1984; and compilation of Wadden
Sea data in Egerrup and Hoegh Laursen 1992). The shell weight to biomass ratio also varies
considerably seasonally (Dankers et al. 1989), as well as from tidal to subtidal (Baird and Drinnan
1957), making mussel carbonate production more difficult to estimate. The bulk of the mussel
population occurs subtidally (Dekker 1989), partly due to mussel culture. For my estimates I have
assumed a biomass (ash-free dry weight) to carbonate ratio of six, based on the average
biomass/length ratio (N. Dankers pers. comm.) and shell weight/length relation measured for
subtidal mussels (Text-figs 8-9). Annual carbonate production by mussels can then be estimated at
6 x 14-7 x 10(i kg. In the western Wadden Sea the average eider population is 48400 (Swennen 1991 ).
Mussel consumption in this part of the Wadden Sea can therefore be estimated at 48400/63000 of
the total consumption of E28xl06kg, and this has to be multiplied by six to arrive at
‘consumption' of mussel carbonate, giving 5-9 x 10Bkg. This equals 15 per cent of the average
annual mussel carbonate production: the same percentage as found for the cockle.

Eiderducks rarely produce shell fragments larger than 8 mm (see Text-fig. 2). They cannot be
responsible for the fragments > 8 mm in bottom samples. In the Wadden Sea bottom samples only
fragments less than 2 mm were examined. To compare shell fragment production by eiderducks
with those present in the bottom samples we therefore have to use fragments present in the 2-8 mm
fraction of our bottom samples and keep in mind that 60 per cent of the fragments in eider faeces
are in this 2-8 mm fraction. Therefore, the eider reduces 10 percent of the annual mussel and cockle
carbonate production to fragments between 2 and 8 mm in size.

In the bottom samples on average 25 per cent of all the shells are whole, 30 per cent are fragments
in the 2-8 mm fraction and 45 per cent are fragments larger than 8 mm (Text-fig. 10). For the cockle
and mussel (Text-figs 1 1-12) respectively 30-4 and 27-9 per cent were found as fragments in the
2-8 mm fraction. This indicates that eider alone could be responsible for c. one-third of the
fragments in this fraction. If we leave out station 3 in the Mokbay, where a relatively high
percentage of fragments was found in the 2-8 mm fraction we can estimate eider contribution at
c. one-half of the 2-8 mm fragments present in bottom samples. However, the fragments in the
Mokbay sample (Text-fig. 13) are very similar to those found in eider faeces (compare with Plate 1,
figures I and 5). As eider roost (and thus also defecate) in flocks on the water, contribution of
eider faeces is not random over the Wadden Sea. This will lead to variation in eider-produced
fragments in bottom samples. The number of six stations may, however, be too small to estimate
accurately the average 2-8 mm fraction of fragments.

OTHER PREDATORS

The large number of fragments larger than 8 mm in the bottom samples, 45 per cent for all molluscs,
38-9 and 45 2 per cent for cockle and mussel respectively (Text-figs 11 12), must be due to factors
other than eider or shelduck predation. Physical factors cannot be ruled out completely. The fact
that the highest number of fragments were found in the intertidal stations (stations 1-3; Text-figs
10-12) suggests wave energy. However, some other predators are also known to produce large
fragments. Oystercatchers feed, like eider, on cockle and mussel in the Wadden Sea (Smit 1981).
They produce characteristic shell fragments, where a small fragment is broken off from one valve

192

PALAEONTOLOGY, VOLUME 37

£

05

5

0

£

W

-o sirtrtidal
intertidal

text-fig. 8. Shell weight (articulated valves) versus shell length for intertidal and subtidal Mytilus, based on
sixty measurements each; only curves given ( r = 098).

mm length

text-fig. 9. Shell length versus ash-free dry weight for subtidal Mytilus , for September and February,
indicating loss of biomass during winter; data obtained from N. Dankers.

only, to enable the bird to sever the adductor muscle (Drinnan 1957, 1958; Hulscher 1964;
Tinbergen and Norton-Griffiths 1964; Tinbergen 1976). The oystercatcher does not always break
a shell, as the bird may succeed in inserting its bill between the valves without damaging them.
Oystercatchers consume almost as many bivalves as the eiderducks in the Wadden Sea (Smit 1981).

CADEE: RECENT SHELL FRAGMENT PRODUCERS

193

□ entire

0 fragm. > 8mm

□ fragm. 2-8mm

text-fig. 10. Distribution of entire mollusc shells, fragments > 8 mm, and fragments from 2-8 mm, in the
> 2 mm fraction of the Wadden Sea bottom samples.

□ entire

□ fragm. > 8mm
13 fragm. 2-8mm

text-fig. 11. Distribution of entire Cerastoderma shells, fragments > 8 mm, and fragments from 2-8 mm, in
the > 2 mm fraction of the Wadden Sea bottom samples.

From Drinnan (1957, 1958) we can estimate that at least 50 per cent of the valves of cockle and
mussel consumed remain intact. Hulscher (1981, 1984) found that oystercatchers feeding on
Macoma balthica left 60-65 per cent of the valves intact.

Another, probably underestimated, predator in the Wadden Sea is the shore crab (Carcinus

194

PALAEONTOLOGY, VOLUME 37

□ entire

(2) fragm. > 8mm
[!§j fragm. 2-8mm

text-fig. 12. Distribution of entire Mytilus shells, fragments > 8 mm, and fragments from 2-8 mm, in the
> 2 mm fraction of the Wadden Sea bottom samples.

CADEE: RECENT SHELL FRAGMENT PRODUCERS

195

maenas). Its omnivorous character is well known; its diet includes bivalves (MacPhail 1955; Ropes
1 968 ; Walne and Dean 1 972 ; LeRoux el al. 1 990). Juvenile crabs feeding on the tidal flats have been
best studied (Klein Breteler 1976). Their annual consumption is comparable to that of all five
species of bird in the Wadden Sea (Swennen 1975). Subtidal feeding of adults on the mussel culture
areas is less well known. Biologists studying the food of this crab have never reported on the
fragments produced, but MacPhail (1955) stated that this crab feeds as a rule on shells as large as
the width of its carapace. Initial studies on shell fragmentation by shore crabs indicate that identical
fragments occur in bottom sediment samples. Finally, shell smashing by herring gulls ( Leans
argentatus) produces large fragments (see Cadee 1989 and references therein), partly by the
dropping of shells on (artificial) hard substrates along the Wadden Sea, but also by dropping them
on the tidal flat.

One-third to one-half of the fragments in the 2-8 mm fraction can be due to eider predation
alone. Herring gulls feeding on molluscs produce, in coughballs and faeces, fragments comparable
in size to those in eider faeces (Schafer 1962; Cadee, unpublished). They are probably as important
as eider in crushing shells in the Wadden Sea (Cadee, unpublished). Smaller bivalves have numerous
other predators in the Wadden Sea: waders, shore crabs, flatfish and shrimps (Reise 1985, fig. 10.4).
This together suggests that most shell fragments in the 2-8 mm fraction are produced by predators.

Entire shells, 25 per cent by weight of the shells in the bottom samples, indicate other causes of
death (see below) or predators that leave the shell intact. This holds for the starfish Asterias rubens ,
which is known as a pest on mussel culture areas, but little quantitative data are available on this
predator in the Wadden Sea. A number of authors have indicated its role as a predator in soft-
bottom communities in NW Europe (Anger et al. 1977; Nauen 1978; Allen 1983). In Kiel Bay
Asterias consumes daily almost 3 per cent of the macrobenthos (including bivalves) according to
Nauen (1978). All sizes of mussels up to the largest (80 mm length) may be consumed by Asterias.
Only specimens with a large adductor muscle may escape Asterias predation (Hancock 1965).
Finally human influence has to be mentioned. On average about 5 per cent of the adult cockle
population is fished annually, but in years with low cockle biomass a higher percentage is fished.
These cockles are cooked at sea, and their empty shells dumped in the Wadden Sea. Mussel fishery
removes shells from the Wadden Sea.

Oystercatchers have already been mentioned which leave > 50 per cent of the valves of bivalves
actually eaten, intact. Cockle mortality in the Wadden Sea, leaving valves intact, may also result
from other factors like parasites, severe winters (Kristensen 1957, 1959), or low oxygen content
(Cadee 1991). Burial of mussels under a layer of storm-deposited sand may also cause mortality
(Kuenen 1942; Theissen 1968). According to Kuenen (1942) mussels covered by more than 20 mm
of sand will die. Mussels may also die in subtidal culture areas when seeded in too thick layers (N.
Dankers pers. comm.).

The large amounts of shells fragmented by predators in the Wadden Sea probably explained most
of the shell fragments found in the Wadden Sea bottom samples. The role of other predators has
yet to be quantified in the same way as done here for eider and shelduck to give a more quantitative
estimate. This awaits the results of further studies now in progress. The high percentage of shell
fragments in Wadden Sea sediments indicates a high predation pressure. However, we have to take
into account the fact that not all predators crush shells. Moreover, LaBarbera (1981) and Walker
and Yamada (1993) observed crushing of empty shells by crabs. This indicates that accurate
predation pressure cannot be estimated from broken shells in a fossil deposit.

PHYSICAL SHELL FRAGMENTATION

Driscoll (1967) was the first to study experimental shell abrasion in the field. He noted that surf
action modifies bivalve shells much more slowly than laboratory abrasion in rolling barrels, which
had been used up to that time to study abrasion. Abrasion starts with removing the surface
sculpture; differential abrasion of various portions of the valve surface may lead to holes in the
shells. Such holes (facets) were described from Wadden Sea shells, particularly Cerastoderma , by

196

PALAEONTOLOGY, VOLUME 37

Pratje (1929) and for the Dutch North Sea coast by Hollmann (1968). A continued abrasion must
lead to shell fragmentation. Driscoll (1967) noted that shell fragmentation was less on sandy beaches
rather than on pebbly beaches as one might expect. On sandy beaches whole valves were relatively
common but few broken valves were present. This indicates that physical shell fragmentation might
occur in exposed areas of the Wadden Sea, but pebbles to assist in shell breaking are scarce.

Cockle shells with abrasion holes near the umbo occur in the Wadden Sea. They can be found
locally in shell concentrations high on exposed tidal flats (e.g. along the Wadden coast of Texel, and
on Janssand, near Spiekeroog, German Wadden Sea, based on my own observations, September
1992). At the same locations fragments of cockle shells occur with clear marks of surface abrasion.
They are not rounded but are sharp edged, and they only differ from fragments produced by
predators by their surface characteristics. However, if fragments produced by predators undergo
abrasion they will look very similar. This makes identification of the process which produced a
particular shell fragment in a bottom sample difficult. Clear abrasion marks (surface sculpture
removed, holes in shells) are rare in the Wadden Sea bottom-samples I studied for this paper, but
only locally in exposed areas this process may predominate. This supports my conclusion that
physical shell fragmentation is of less importance than biological fragmentation in the Wadden Sea
as a whole. This observation was based on the fact that quantitative data available on predators and
their consumption and fragmentation of shells can account for a large part of the fragments found
in the sediment.

Linke (1939), Van Straaten (1954), and Dorjes et al. (1969) observed relatively high amounts of
carbonates in the fine grainsize fractions of the Wadden Sea sediments. Reineck (1970, p. 32)
explains this by physical destruction (abrasion) of shells. Results presented here indicate that also
eider (PI. 1, figs 4, 8; Text-figs 2-3) and probably other shell-crushing predators, produce fine
carbonate particles. Shell-boring organisms will also add fine-grained carbonate particles to the
sediment.

PALAEOECOLOGICAL IMPLICATIONS

Is the present the key to the past? Was shell fragmentation by predators in comparable habitats in
the past as important as it is now in the Wadden Sea? Shell-breaking predators are known to have
existed in the Early Cambrian. A large-scale diversification of such predators occurred in the
Devonian and a large increase in families of specialized shell-breaking predators took place during
the Late Cretaceous and Early Cenozoic (Vermeij 1987). Ducks are a late addition to this guild of
shell-breaking predators. As for the predators described here, Anseriformes are known since the
Eocene, fossil eider are only known from the Pleistocene (Uspenski 1972), but as the fossilization
potential of birds is low (Schafer 1962), they may have appeared earlier.

The present-day high intensity of shell-breaking by predators is probably characteristic for most
of the Cenozoic. However, this intensity of shell fragmentation by predators was probably less
before the Late Cretaceous, although the Early Palaeozoic shell fragments have been produced by
predators. Palaeoecologists should be aware of the role of shell-breaking predators in producing
shell fragments in marine environments. Such fragments may be of all sizes, and usually will be
sharp edged, sometimes characteristic for a certain predator. In most cases they may be comparable
to fragments produced by physical processes, particularly if they become abraded subsequently.
'Facets’, holes in shells produced by abrasion, are the best indicators of fragmentation by physical
processes, since such holes cannot be produced by predators. According to Pratje (1929) facets are
related to the action of (tidal) currents confined to intertidal and shallow subtidal areas.

HIGH FIDELITY

Kidwell and Bosence (1991) suggested methods to test the fidelity of the death assemblage to the
live shelly fauna. Our death assemblage data are based on weight percentages, while the data for
the live shelly fauna are based on species lists and biomass data, and not for the same stations but
for the entire Dutch Wadden Sea. However, live fauna data are available for a period of twenty-

CADEE: RECENT SHELL FRAGMENT PRODUCERS

197

five years for the tidal flats (Beukema 1976, 1981, 19826, 1989) and for a shorter period in the
snbtidal parts of the Wadden Sea (Dekker 1989). Of Kidwell and Bosence’s tests two are applicable:
the percentage of dead species that are also found alive, and the comparison of the rank order of
species alive and in the death assemblage (Table 3).

table 3. Comparison of rank order of species in death assemblage and live shelly fauna.

Overall Death assemblage

Intertidal biomass Subtidal biomass rank > 2 mm

(Beukema 1981) (Dekker 1989) living

Species

g afdw

(rank)

g afdw

(rank)

— shelly
fauna

per cent shell
weight (rank)

Mvtiius edulis

6-2

(1)

28-7

(1)

1

44-0

(2)

Cerastoderma edide

4-3

(2)

14

(3)

2

445

(1)

Mva arenaria

4-2

(3)

04

(5)

4

2-4

(4)

Macoma baltliica

2-2

(4)

14

(4)

5

2-3

(5)

Hydrobia ulvae

0-2

(5)

49

(2)

3

< 0-1

—

Petricola pholadiformis

—

—

0-6

(6)

6

0-2

—

Ensis directus

—

—

—

—

—

0-9

(6)

Littorina littorea

—

—

—

—

—

3-5

(3)

Few species in our bottom samples were not found alive in the area. I have already referred to
the Eemian Bittium reticulatum. Two other small gastropods, Rissoa membranacea and Lacuna
vincta , found in very small numbers in some of the samples, were formerly common in the seagrass
( Zostera ) meadows (Van Benthem Jutting 1933), but disappeared after the ‘wasting disease’
destroyed the subtidal Zostera vegetation in the Wadden Sea in 1932 (Den Hartog 1987). Ostrea
edulis also has disappeared from the Dutch Wadden Sea (Wolff and Dankers, 1981). The small
bivalve Saxicave/la jeffreysi , of which we found one valve, has never been reported alive in Dutch
coastal waters (Van Benthem Jutting 1943). Thus of the twenty-six species found dead, one has
disappeared since the Eemian, two since 1932, Ostrea this century and one has never been found
alive. This results in a fidelity percentage of 80 per cent or even 92 per cent if we include data on
the living fauna of the first half of this century. Kidwell and Bosence (1990, 1991) reported lower
fidelity percentages (31-49 per cent) in their review of available data, when the live community was
censused only once. Fidelity percentages climbed to 70-80 per cent when live census data were
pooled over successive years or decades. The Wadden Sea percentage is even higher, probably
because the living fauna is longer studied and poor in species. Also if we compare the rank of species
we arrive at a high fidelity of the death assemblage in the live shelly fauna. For the live fauna
( > 1 mm) we combined data of Beukema (1981 ) and Dekker (1989) to get an overall rank of species
(Table 3). The first and second place in both is for Mytilus and Cerastoderma , albeit in different
order, and this can partly be explained by a higher shell carbonate/biomass ratio in Cerastoderma.
Hydrobia , third in the living fauna ( > I mm) was too small to reach a high rank in the > 2 mm
fraction. Mya and Macoma in both live and dead faunas occupy positions four and five. Number
three in the dead fauna Littorina may be overestimated in the death assemblage as the samples were
taken relatively close to the coast where this species occurs in higher numbers than on the tidal flats.
Ensis directus is a recent American immigrant in the Wadden Sea. It appeared in 1979, in the
German Bight (Von Cosel et al. 1982), and reached the Dutch Wadden Sea in 1982 (Essink 1985).
It is therefore not included in the data of Beukema (1981), and Dekker (1989) dealt with pr e-Ensis
directus data (1981/82). In the 1992 samples of empty shells it already takes sixth place.

198

PALAEONTOLOGY, VOLUME 37

CONCLUSION

The results from this study indicate the importance of predators in shell fragmentation. This is
not only be the case in shallow coastal seas where birds are the main predators in the intertidal zone.
In deeper waters shell crushing by fish and crustaceans will be important. The results cast doubt on
the use of a crush factor as an indicator of the degree of wear of the shell material during transport
(Van Straaten 1956; Ager 1964). Correlations between fragmented shells with only water turbulence
as suggested by, for example, Vokes (1948), Link (1967), Bissell and Chillinger (1967) do not seem
possible. A fragmentation rate of shells to assess the autochthonous/allochthonous ratio of
assemblages as suggested by Sato and Shimoyama (1992) does not take into account the role of
predators. This biological fragmentation does not influence fidelity of the death assemblage to the
live fauna. The high fidelity of the Wadden Sea molluscan death assemblage is encouraging for
palaeoecologists. If physical abrasion had been more important, only durable skeletons would have
been preserved and the fidelity to the live shelly fauna would have been low (Chave 1964). This will
occur in the exposed surfzone of the open ocean, but not in the relatively sheltered Wadden Sea.

Acknowledgements. I am very grateful to Dr N. Dankers (Institute for Forestry and Nature Research, Texel)
for providing me with mussel meat versus shell length data (Text-fig. 9); for critical remarks improving the
typescript I thank J. J. Beukema, N. Dankers, K. W. Flessa, S. M. Kidwell, T. Piersma, C. Swennen and the
two referees.

REFERENCES

ager, D. v. 1962. Principles of palaeoecology. MacGraw-EIill, New York, ix + 371 pp.
allen, p. l. 1983. Feeding behavior of Asterias rubens (L.) on soft bottom bivalves: a study in selective
predation. Journal for Experimental Marine Biology and Ecology , 70, 79-90.
anger, K., rogal, u., schriever, G. and valentine, c. 1977. In-situ investigations on the echinoderm Asterias
rubens, as a predator of soft-bottom communities in the western Baltic Sea. Helgolander wissenschaftliche
Meeresuntersuchungen, 29, 439-459.

baird, r. h. and drinnan, r. e. 1957. The ratio of shell to meat in Mytilus as a function of tidal-exposure to
air. Journal du Conseil International de /’ Exploration de la Mer, 22, 329-336.
bauer, K. m. and glutz VON blotzheiim, E. n. 1968. Handbiich der Vogel Mitteleuropas. Band 2. Akademische
Verlagsgesellschaft, Frankfurt-am-Main, 535 pp.

beukema, j. j. 1976. Biomass and species richness of the macrobenthic animals living on the tidal flats of the
Dutch Wadden Sea. Netherlands Journal of Sea Research, 10, 236—261.

— 1980. Calcimass and carbonate production by molluscs on the tidal flats in the Dutch Wadden Sea. I. The
telhnid bivalve Macoma balthica. Netherlands Journal of Sea Research , 14, 323-338.

1981. Quantitative data on the benthos of the Wadden Sea proper. 134-142. In wolff, w. j. (ed. ).
Invertebrates of the Wadden Sea. Reports of the Wadden Sea working group, 4. Balkema, Rotterdam,

221 pp.

1982a. Calcimass and carbonate production by molluscs on the tidal flats in the Dutch Wadden Sea. II.
The edible cockle, Cerastoderma edule. Netherlands Journal of Sea Research, 15, 391 — 405.

19826. Annual variation in reproductive success and biomass of the major macrozoobenthic species living
in a tidal flat area of the Wadden Sea. Netherlands Journal of Sea Research, 16,

— 1989. Long-term changes in macrozoobenthic abundance on the tidal flats of the western part of the
Dutch Wadden Sea. Helgoldnder Wissenschaftliche Meeresuntersuchungen, 43, 405-415.

bryant, D. M. and leng, J. 1975. Feeding distribution and behaviour of Shelduck in relation to food supply.
Wildfowl, 26, 20-30.

cadee, G. c. 1968. Molluscan biocoenoses and thanatocoenoses in the Ria de Arosa, Galicia, Spain.
Zoologische Verhandelingen Leiden, 95, 1-121.

— 1988. Levende wadslakjes in Bergeend faeces. Correspondentieblad van de Nederlandse Malacologische
Vereniging , 243/4, 443-144.

1989. Size selective transport of shells by birds and its palaeoecological implications. Palaeontology, 32,
429-437.

— 1990a. Feeding traces and bioturbation by birds on a tidal flat, Dutch Wadden Sea. Ichnos, 1, 23-30.

— 19906. Lokale sterfte van kokkels op het wad tijdens een Noctiluca bloei. Het Zeepaard, 50, 119-128.

CADEE: RECENT SHELL FRAGMENT PRODUCERS

199

— 1991. De alikruik (Littorina littorea ) als eidereend voedsel. Correspondentieblad van de Nederlandse
Malacologische Vereniging , 262, 876-880.

— and hegeman, J. 1975. Levende afgeknotte strandschelpen in Waddenzee. De Levende Natuur , 78, 48.
carter, r. m. 1968. On the biology and palaeontology of some predators of bivalved mollusca.

Palaeogeography , Palaeo climatology, Palaeoecology , 4, 29-65.
chave, K. E. I960. Carbonate skeletons to limestones: problems. Transactions of the New York Academy of
Sciences , 23, 14-24.

— 1964. Skeletal durability and preservation. 377-387. In imbrie, j. and Newell, n. d. (eds). Approaches to
palaeoecology. Wiley, New York, viii + 432 pp.

dankers, n., koelemaij, k. and zegers, j. 1989. De rol van de mossel en de mosselcultuur in het ecosysteem
van de Waddenzee. Rijks Instituut voor Natuurbeheer , Rapport , 89/9, 1-66.
dekinga, a. and piersma, t. 1993. Reconstructing diet composition on the basis of faeces in a mollusc-eating
wader, the Knot Calidris canutus. Bird Study , 40. 144- 156.
dekker, R. 1979. Numbers, growth, biomass and production of organic and calcareous matter of Hydrobia
ulvae (Gastropoda: Prosobranchia) in the western Dutch Wadden Sea. Interne Verslagen Nederlands
Instituut voor Onderzoek der Zee , Texel, 1979-15, 1-27.

1989. The macrozoobenthos of the subtidal western Dutch Wadden Sea. I. Biomass and species richness.

Netherlands Journal of Sea Research , 23, 57-68.

den hartog, c. 1987. ‘Wasting disease’ and other dynamic phenomena on Zostera beds. Aquatic Botany , 27,
3-14.

dories, j., gadow, s., reineck, H.-E. and singh, i. b. 1969. Die Rinnen der Jade (Sudliche Nordsee). Sedimente
und Makrobenthos. Senckenbergiana Maritime, 1, 5-62.
drinnan, r. e. 1957. The winterfeeding of the oystercatcher (Haematopus ostralegus) on the edible cockle
(Cardium edule). Journal of Animal Ecology , 26, 441-469.

1958. The winterfeeding of the oystercatcher (Haematopus ostralegus) on the edible mussel (Mytilus

edulis ) in the Conway estuary. North Wales. Fishery Investigations, Series 2, 22, 1-15.
driscoll, E. G. 1967. Experimental field study of shell abrasion. Journal of Sedimentary Petrology , 37,
1117-1123.

egerrup, m. and hoegh laursen, m.-l. 1992. Aspects of predation on intertidal blue mussels (Mytilus edulis
L.) in the Danish Wadden Sea. International Council for the Exploration of the Sea , Shellfish Committee ,

C.M.1992/K : 25.

eisma, D. 1966. The distribution of benthic marine molluscs off the main Dutch coast. Netherlands Journal of
Sea Research , 3, 107-163.

— and hooft, j. j. 1969. The formation of shell fragments in the southern North Sea. Netherlands Institute
for Sea Research , Publications and Reports , 1969-3, 1-17.

essink, K. 1985. On the occurrence of the American jack-knife clam Ensis directus (Conrad, 1843) (Bivalvia,
Cultellidae) in the Dutch Wadden Sea. Basteria , 49, 73-79.
feige, a. and fursich, f. t. 1991. Taphonomy of the Recent molluscs of Bahia la Choya (Gulf of California,
Sonora, Mexico). Zitteliana , 18, 89-133.

flemming, b. w., Schubert, h., hertweck, G. and muller, k. 1992. Bioclastic tidal-channel lag deposits: a
genetic model. Senckenbergiana Maritime i, 22, 109-129.
force, l. m. 1969. Calcium carbonate size distribution on the West Florida shelf and experimental studies on
the microarchitectural control of skeletal breakdown. Journal of Sedimentary Petrology, 39, 902-934.
ginsburg, R. n. 1957. Early diagenesis and lithification of shallow- water carbonate sediments in South Florida.
80-100. In leblanc, r. j. and breeding, j. g. (eds). Regional aspects of carbonate deposition. Society of
Economic Paleontologists , Special Publication, 5, 178 pp.

Hancock, d. a. 1965. Adductor muscle size in Danish and British mussels and its relation to starfish predation.
Ophelia, 2, 253-267.

hollmann, r. 1968. Zur Morphologie rezenter Mollusken-Bruchschille. Paldontologisches Zeitschrift , 42,
217-235.

hulscher, j. B. 1964. Scholeksters en lamellibranchiaten in de Waddenzee. De Levende Natuur, 67, 80-85.

— 1975. Het wad, een overvloedig of schaars gedekte tafel voor vogels? 57-82. In swennen, c., de wilde,
p. a. w. J. and haeck, J. (eds). Symposium Waddenonderzoek , Mededelingen Werkgroep W addengebied, 1,
93 pp.

1982. The oystercatcher as a predator of Macoma. Ardea , 70, 80-152.
kidwell, s. m. and baumiller, t. 1990. Experimental disintegration of regular eehmoids: roles of temperature,
oxygen, and decay thresholds. Paleobiology , 16, 247-271.

200

PALAEONTOLOGY, VOLUME 37

kidwell, s. m. and bosence, d. w. j. 1990. Quantitative fidelity of time-averaged death assemblages: good news
for paleobiology. Geological Society of America, Abstracts with Program , 22, A83.

- 1991. Taphonomy and time-averaging of marine shelly faunas. 115-209. In allison, p. a. and
briggs, D. E. (eds). Taphonomy : releasing the data locked in the fossil record. Plenum Press, New York,
xiv + 560 pp.

klahn, h. 1932. Der quantitative Verlauf der Aufarbeitung von Sanden, Gerollen und Schalen in wasserigem
Medium. Neues Jahrbuch fur Mineralogie, Geologie und Palaontologie , 67 (B) (Supplement), 313-412.
klein breteler, w. c. m. 1976. Settlement, growth and production of the Shore Crab, Carcinus maenas, on
tidal flats in the Dutch Wadden Sea. Netherlands Journal of Sea Research , 10, 354-367.
kotler, e., martin, r. e. and liddell, w. d. 1992. Experimental analysis of abrasion and dissolution resistance
of modern reef-dwelling Foraminifera : implications for the preservation of biogenic carbonate. Palaios , 7,
244-276.

krause, h. r. 1950. Quantitative Schilluntersuchungen im See- und Wattengebiet von Noderney und Juist und
ihre Verwendung zur Klarung hydrographischer Fragen. Archiv fur Molluskenkunde , 79, 91-116.
kristensen, i. 1957. Differences in density and growth in a cockle population in the Dutch Wadden Sea.
Archives Neerlandaises de Zoologie , 12, 351 — 453.

- 1959. The coastal waters of the Netherlands as an environment of molluscan life. Basteria , 23
(Supplement), 18-55.

kuenen, D. J. 1942. On the distribution of mussels on the intertidal sand flats near Den Fielder. Archives
Neerlandaises de Zoologie , 4, 117-160.

labarbera, m. 1981. The ecology of Mesozoic Gryphaea, Exogyra , and Ilymatogyra. Paleobiology , 7, 510-526.
le roux, p. j., branch, G. M. and joska, m. a. p. 1990. On the distribution, diet and possible impact of the
invasive European Shore crab Carcinus maenas (L.) along the South African coast. South African Journal
of Marine Science , 9. 85-93.

link, a. g. 1967. Delineating the major depositional environments in northern Port Philip Bay, Victoria.
Journal of Sedimentary Petrology , 37, 931-951.

linke, o. 1939. Die Biota des Jadebusenwattes. Helgolander Wissenschaftliche Meeresuntersuchungen , 1,
201-348.

macphail, j. s. 1955. The green crab: a new clam enemy. Atlantic Progress Reports, 63, 3-12.
nauen, c. e. 1978. The growth of the sea star, Asterias rubens, and its role as benthic predator in Kiel Bay.
Kieler Meeresforschungen, 4, 68-81.

Norton, s. f. 1988. Role of the gastropod shell and operculum in inhibiting predation by fishes. Science, 241,
92-94.

offringa, h. 1991. Verspreiding en voedseloecologie van de Zwarte Zeeeend (Melanitta nigra ) in Nederland.

Nederlands Instituut voor Onderzoek der Zee, Rapport, 1991-13, 1-60.
palmer, d. c. 1977. A palaeontologist’s view of shelly invertebrates. 49-66. In Jeffrey, d. w. (ed.). North Bull
Island, Dublin Bay - a modern coastal natural history. Royal Dublin Society, Dublin, 158 pp.
parsons, k. m. and brett, c. e. 1991. Taphonomic processes and biases in modern marine. In
donovan, s. k. (ed.). The processes of fossilization. Belhaven Press, London, xi + 303 pp.
piersma, t., koolhaas, a. and dekinga, a. 1994. Interaction between stomach structure and diet choice in shore
birds. Auk, in press.

pilkey, o. h. 1964. The size distribution and mineralogy of the carbonate fraction of the United States South
Atlantic shelf and upper slope sediments. Marine Geology , 2, 121-136.

— backwelder, B. w., doyle, l. j., estes, e. and terlecky, p. m. 1969. Aspects of carbonate sedimentation
of the Atlantic continental shelf off the Southern United States. Journal of Sedimentary Petrology, 39,
744-768.

powell, e. n., staff, g. m., Davies, d. j. and callendar, w. r. 1989. Macrobenthic death assemblages in
modern marine environments: formation, interpretation, and application. Critical Reviews in Aquatic
Sciences, 1, 555-589.

pratje, o. 1929. Fazettieren von Molluskenschalen. Palaontologische Zeitschrift , 11, 151-169.

reineck, h.-e. 1970. Das Watt Ablagerungs- und Lebensraum. Waldemar Kramer, Frankfurt-am-Main, 142 pp.

reise, K. 1985. Tided flat ecology. Springer Verlag, Berlin, vii+191 pp.

ropes, j. w. 1968. The feeding habits of the green crab, Carcinus maenas (L.). Fishery Bulletin, 67, 183-203.
sato, s. and shimoyama, s. 1992. Paleoenvironmental analysis of the middle part of the Shimosa Group on the
basis of pelecypod fossil assemblages. Journal of the Geological Society of Japan, 98, 529-545.
schafer, w. 1962. Aktuo-paldontologie nach Studien in der Nordsee. Waldemar Kramer, Frankfurt-am-Main,
viii + 666 pp.

CADEE: RECENT SHELL FRAGMENT PRODUCERS

201

sha, l. p. 1989. Holocene- Pleistocene interface and three-dimensional geometry of the ebb-delta complex,
Texel inlet, the Netherlands. Marine Geology , 89. 207-228.
smit, c. j. 1981. Production of biomass by invertebrates and consumption by birds in the Dutch Wadden Sea
area. 290-301. In smit, c. j. and wolff, w. j. (eds). Birds of the Wadden Sea. Reports of the Wadden Sea
working group , 6. Balkema, Rotterdam, 308 pp.

spaink, G. 1965. De Nederlandse Eemlagen. Wetenschappelijke Mededelingen Koninklijke Nederlandse
Natuurhistorische Vereniging , 29, 1-44.

staff, G. m. and powell, e. n. 1990. Local variability of taphonomic attributes in a parautochthonous
assemblage: can taphonomic signature distinguish a heterogeneous environment? Journal of Paleontology ,
64, 648-658.

swennen, c. 1975. Aspecten van voedselproduktie in de Waddenzee en aangrenzende zeegebieden in relatie met
de vogelrijkdom. Het Vogeljaar , 23, 141-156.

— 1976. Populatie-structuur en voedsel van de eidereend Somateria m. mollissima in de Nederlandse
Waddenzee. Ardea , 64, 311-371.

— 1991. Ecology and population dynamics of the common eider in the Dutch Wadden Sea. Unpublished
thesis. University of Groningen.

swinchatt, j. p. 1965. Significance of constituent composition, texture, and skeletal breakdown in some recent
carbonate sediments. Journal of Sedimentary Petrology , 35, 71-90.
theissen, x. 1968. Growth and mortality of culture mussels in the Danish Wadden Sea. Meddelelser fra
Danmarks Fiskeri- og Havundersogelser , Ny Serie 6, 47-78.

Thompson, d. b. a. 1982. The abundance and distribution of intertidal invertebrates, and an estimation of their
selection by Shelduck. Wildfowl , 33, 151-158.

tinbergen, n. and norton-griffiths, m. 1964. Oystercatchers and mussels. British Birds , 57, 64—70.
1976. De mosseleters van Ravensglass. De Levende Natuur , 79, 1-14.

trewin, n. h. and welsh, w. 1972. Transport, breakage and sorting of the bivalve Mactra corallina on
Aberdeen Beach, Scotland. Palaeogeography , Palaeoclimatology , Palaeoecology , 12, 193-204.

1976. Formation and composition of a graded estuarine shell bed. Palaeogeography ,
Palaeoclimatology , Palaeoecology , 19, 219-230.

uspenski, s. m. 1972. Die Eiderenten. Die Neue Brehm-Biicherei, 452. A. Ziemsen Verlag, Wittenberg
Lutherstadt, 103 pp.

van benthem jutting, t. 1993. Fauna van Nederland. VII. Molluscct Gastropoda Prosobranchia et Pulmonata.
Sijthoff, Leiden, 387 pp.

— 1943. Fauna van Nederland. XII. Mollusca Lamellibranchia. Sijthoff, Leiden, 477 pp.

van regteren altena, c. o. 1937. Bijdrage tot de kennis der fossiele, subfossiele en recente mollusken, die op
de Nederlandse stranden aanspoelen, en hunner verspreiding. Nieuwe Verhandelingen van het Bataafs
Genootschap te Rotterdam (2), 10, 1-184.

— bloklander, a. and pouderoyen, l. p. 1955. De fossiele schelpen van de Nederlandse stranden en
zeegaten. 2. Basteria , 29, 27-34.

van straaten, l. M. j. u. 1952. Biogene textures and the formation of shell beds in the Dutch
Wadden Sea. Proceeding van de Koninklijke Nederlandse Akademie van Wetenschappen , Series B , 55,
500-516.

— 1954. Composition and structure of recent marine sediments in The Netherlands. Leidse Geologische
Mededelingen , 19. 1-108.

— 1956. Composition of shellbeds formed in tidal flat environment in The Netherlands and in the Bay of
Arcachon (France). Geologie en Mijnbouw , 18, 209-226.

VERMEH, G. J. 1978. Biogeography and adaptation , patterns of marine life. Harvard University Press, Cambridge,
332 pp.

— 1987. Evolution and escalation an ecological history of life. Princeton University Press, Princeton,
xv + 527 pp.

verrill, a. e. 1882. Notice of the remarkable marine fauna occupying the outer banks off the southern coast
of New England, number 8. Nature and origins of the sediments. Occurrence of fossiliferous limestone-
nodules. American Journal of Science , 24, 447 — 452.

vokes, H. E. 1948. Cretaceous mollusca from depths of 4875 to 4885 feet in the Maryland Esso well. Maryland
Department of Geology , Mines and Water Research , Bulletin , 2, 126-151.
von cosel, r., dorjes, j. and muhlenhardt-siegel, u. 1982. Die Amerikanische Schwertmuschel Ensis directus
(Conrad) in der Deutschen Bucht. 1. Zoogeographie und Taxonomie im Vergleich mit den einheimischen
Schwertmuschel-Arten. Senckenbergiana Maritima , 14. 147-173.

202

PALAEONTOLOGY, VOLUME 37

walker, s. E. and yamada, s. B. 1993. Implications for the gastropod fossil record of mistaken crab predation
on empty mollusc shells. Palaeontology , 36, 735-741.

walne, p. r. and dean, G. j. 1972. Experiments on predation by the shore crab, Carcinus maenas L. on
Mytilus and Mercenaria. Journal du Conseil International de t Exploration de la Mer , 34, 190-199.
walther, j. 1893. Lithogenesis der Gegenwart. Einleitung in die Geologie als historische Wissenschaft, 3,
535-1055. Gustav Fischer, Jena.

— 1910. Die Sedimente der Taubenbank im Golfe von Neapel. Abhandlungen der koniglich Preussischen
Akademie der Wissenschafte , Philosophisch- Historische Classe, 1910(3), 149.
wolff, w. j. and dankers, n. 1981. Preliminary checklist of the zoobenthos and nekton species of the Wadden
Sea. 24-60. In dankers, n., kuhl, h. and wolff, w. j. (eds). Invertebrates of the Wadden Sea. Report of the
Wadden Sea Working Group , 4, 221 pp.

woodward, s. p. 1875. A manual of the Mollusca. Lockwood, London, xiv + 542 pp.

GERHARD C. CADEE

Netherlands Institute for Sea Research
PO Box 59

Typescript received 13 October 1992 1790 AB Den Burg, Texel

Revised typescript received 10 March 1993 The Netherlands

THE LEPTOCERATOIDINAE: SMALL
HETEROMORPH AMMONITES FROM THE

BARREMI AN

by ZDENEK VASICEK and JOST WIEDMANNt
t deceased 2 December 1993

Abstract. The Barremian Leptoceratoidinae, a subfamily of small-sized heteromorph ammonites, are revised.
Despite considerable differences in the mode of uncoiling, Leptoceratoidinae are considered to represent a
monophyletic unit, defined by their small size, simplified suture-lines and ubiquitous Barremian age. Three
evolutionary lines are recognized exhibiting parallel trends of size increase: (1) Karsleniceras with criocone
uncoiling and a planispiral to trochospiral initial coil ; (2) Hamulinites with an ancylocone type of uncoiling and
a small planispiral initial coil ; and (3) a line compromising Eoheteroceras gen. nov. with ancylocone uncoiling
and a large trochospiral initial coil, as well as Manoloviceras gen. nov. with only one slightly curved arm and
a large trochospiral initial coil. It is inferred that these evolutionary lines originated in Veveysiceras gen. nov.
with hamitid-like uncoiling. Veleziceras, with a straight to gently curved shell, is tentatively included in the
Leptoceratoidinae, but cannot yet be assigned to one of the three lines defined above. The following new species
are described : Karsteniceras ibericum, K. beyrichoide , K. hoheneggeri , K. trinidadense and E. silesiacum , the type
species of Eoheteroceras gen. nov. The origin of leptoceratoids remains obscure; they are not closely related
to the Berriasian to Valanginian true leptoceratids. Presumably, Eoheteroceras gave rise to the large-sized
Heteroceras. Due to the simplified suture-line formula ELUI and the type of uncoiling, the Leptoceratoidinae
are included in the Ancyloceratidae. Their evolutionary centre was the southern (Tethyan) margin of the
European Plate. From this area, Leptoceratoidinae migrated into the central North Atlantic and even the
western Pacific. While Leptoceratoidinae presumably had a vagrant epibenthic mode of life, their dispersal was
most probably achieved during their nearshore juvenile stage by the North Equatorial Current.

When Uhlig (1883) created Leptoceras as a subgenus of Crioceratites, a long history of
misinterpretation, misconception and misdating began. Uhlig was convinced that these small-sized
criocones were a well defined monophyletic group of Barremian age. More than sixty years later,
Thieuloy (1966) demonstrated convincingly that Uhlig lumped at least two different but
homoeomorphic groups together, one of Berriasian to Valanginian, the other of Barremian age.
Unfortunately, Uhlig’s type species, Leptoceras brunneri, belongs to the earlier stock. Both groups
are phylogenetically separate, they exhibit different suture patterns, and no transitional forms of
Hauterivian age are found. Since, however, the Barremian stock comprises criocones, ancylocones
and hamiticones with planispiral or trochospiral initial coils, the question arises whether all these
forms are really monophyletic or micromorphic descendants of these different groups. This was the
starting point of the present revision. Difficulties were encountered because some of the type
specimens are lost, and others too poorly preserved to allow reliable identification and
interpretation. Therefore, recollection of material became necessary to obtain well-preserved
specimens from well-dated sections.

HISTORICAL REVIEW

The extremely difficult and controversial interpretation of Leptoceratoidinae can be evaluated only
by reviewing the history of research. The first scanty representatives were described by d’Orbigny
(1850) from France, by Karsten (1858) from Colombia, and by Ooster (1860) from Switzerland, as
small Crioceratites or Ancyloceras. Uhlig (1883) realized the peculiarity of these forms when he

| Palaeontology, Vol. 37, Part I, 1994, pp. 203-239, 4 pls.|

© The Palaeontological Association

204

PALAEONTOLOGY, VOLUME 37

studied a relatively rich fauna from the Barremian of Silesia. Based on the simplified suture-lines,
he considered the ancylocone and criocone 'microconchs' to represent a new subgenus Leptoceras
of Crioceratites. Unfortunately, no type species was designated. Leptoceras was accepted by Sayn
(1891) for small fragments of criocones (L. cirtae ) first mentioned by Conquand (1880) from the
Barremian of Algeria. Nickles (1894) described a small Barremian hamulinicone from the Subbetic
of Spain which was designated the type of the new genus Hamulinites by Paquier (1901). Sarasin and
Schondelmayer (1902), when redescribing Lower Cretaceous ammonites from Switzerland, accepted
some of Ooster’s species but considered them to be juvenile fragments of well-known large-sized
genera.

New species of Leptoceras were described by Yabe et at. (1926) from Japan, and by Stahlecker
( 1935) from the Barremian of Maio, Cape Verde Islands. The Essai de genera published by Roman
(1938) became the source of serious problems for the next few decades: Roman proposed
Ancyloceras brunneri Ooster, the first listed species of Uhlig, as the type species, but figured
erroneously Leptoceras pumilum Uhlig under the name L. parvulum Uhlig (Roman 1938, pi. 35, figs
335-336). Much later, Thieuloy (1966) realized that, beginning with Ooster (1860), two similar but
distinct groups of ‘microconchs’ had been included in Leptoceras : (1) a Berriasian group centred
around L. brunneri (Ooster); and (2) a Barremian group centred around L. pumilum Uhlig.

Meanwhile, Royo y Gomez (1945) proposed the new genus Karsteniceras for Ancyloceras beyrichi
Karsten, 1858, and also created the similar genus Orbignyceras with O. veleziense as type species.
Both genera were first described from the Barremian of Colombia.

Wright (1957) renamed the pre-occupied Orbignyceras as Veleziceras , and included it together
with Karsteniceras in his subfamily Crioceratitinae, while Leptoceras , with L. pumilum Uhlig as type
species, was considered to belong to Ancyloceratinae. This obvious misinterpretation of the
criocone L. pumilum may be due to Roman's previous mistake. Luppov and Drushtchic (1958)
followed the same interpretation.

Manolov ( 1962) created the new genus Eoleptoceras for the ancylocone species L. parvulum Uhlig,
with two more new subgenera, Tzankoviceras and Wrighticeras. These and the criocones Leptoceras ,
Karsteniceras and Veleziceras were all considered to be of Barremian age and to form the new
crioceratid subfamily Leptoceratinae.

Wiedmann (1963) drew attention to the fact that Eoleptoceras and Tzankoviceras Manolov were
objective synonyms, and that they were subjective synonyms of Hamulinites Paquier and Wrightites
Manolov. He also described the suture ontogeny of Hamulinites.

When Thieuloy (1966) separated the Berriasian stock from that of the Barremian, he created the
new genus Leptoceratoides for the Barremian forms around Leptoceras pumilum Uhlig, and
proposed, accordingly, to restrict Leptoceratinae to the early stock and created Leptoceratoidinae
for the Barremian forms. Leptoceras Uhlig (type species L. brunneri ), however, was restricted to
the Berriasian, and included in the Protancyloceratinae Breistroffer and consequently, in the
Bochianitidae Spath. Nikolov (1966) described a new genus Protoleptoceras , based on P. jelevi
Nikolov, from the Berriasian of Bulgaria, which became a synonym of Leptoceras.

Breskovski (1966), Dimitrova (1967), and Nagy (1967) added further material from Bulgaria and
Hungary, following Wright's and Manolov’s systematic concepts. Dimitrova included, moreover,
the Berriasian Protoleptoceras Nikolov in 'Leptoceratinae' sensu Manolov. In 1968, Etayo Serna
revised the Colombian Karsteniceras beyrichi (Karsten) and placed Karsteniceras in the
Leptoceratoidinae Thieuloy.

In 1970, Dimitrova made an attempt to classify Lower Cretaceous heteromorphs based on their
shape and adult suture-lines. The result is, however, hard to understand since the Berriasian
Protoleptoceras was lumped with the Barremian Karsteniceras in the Scaphitaceae which
originated in the Albian. All the remaining genera ( Leptoceras , Tzankoviceras , Eoleptoceras ,
Hamulinites) were considered Barremian, and again included in the Leptoceratidae and these in the
Protancylocerataceae.

Vasicek ( 1972) followed Thieuloy’s classification when describing new discoveries of leptoceratids
from the Moravo-Silesian Barremian. Wiedmann (1973) also followed Thieuloy, at least in

VASICEK AND WIEDMANN: BARREMI AN AMMONITES

205

including the Berriasian Leptoceras in Protancyloceratinae; the Barremian microconchs were
reduced to two genera, Karsteniceras Royo y Gomez ( = Veleziceras , = Leptoceratoides ) for the
criocones, and Hamulinites Paquier (= Eoleptoceras , = Tzankoviceras, — Wrightites) for the
ancylocones, and these two genera were included in the Ancyloceratinae Meek. Myczynski (1977)
and Myczynski and Triff (1986) followed this interpretation when describing species of Karsteniceras
and Hamulinites from the Barremian of Cuba. Roumanian leptoceratoids were described by Avram
(1976) who used Leptoceratidae Manolov and again included Karsteniceras in the Scaphitidae
(sensu Dimitrova 1970).

New species were recorded and described from the Southern Alps (Rieber 1977), northern
Calcareous Alps (Darga and Weidich 1986; Immel 1987) and, again, from Roumania (Avram and
Kusko 1984). Immel (1987) preferred to attribute Karsteniceras to the Crioceratitinae, while Wright
(1981) synonymized Leptoceratoidinae with the Helicancylinae. This idea was shared by Gonzalez-
Arreola and Carillo-Martinez (1987) when they published on Mexican Karsteniceras. Finally,
Matsukawa (1987) followed Wiedmann’s classification in his careful study of the early ontogeny of
a new Japanese Karsteniceras.

Despite the fact that the rather divergent views of the 1960s have become more focused during
the last decade, the systematics of this group is still unsatisfactory and needs revision.

SYSTEMATIC PALAEONTOLOGY

Material. The material studied is deposited in the following collections: BSM, Bayerische Staatssammlung fur
Palaontologie und Historische Geologie, Miinchen; GBAW, Geologische Bundesanstalt, Wien; GPIT,
Geologisch-Palaontologisches Institut, Tubingen; MLING, Musee d’Histoire Naturelle, Geneve; MHNP,
Musee d’Histoire Naturelle, Paris; NMB, Naturhistorisches Museum, Bern; OSM, Ostrava-Museum,
Ostrava; SNM, Slovak National Museum, Bratislava; USNM, National Museum of Natural Elistory,
Smithsonian Institution, Washington; VSB, Vysoka Skola Banska, Ostrava.

Additional unstudied type material is kept in the following collections: HNM, Hungarian National
Museum, Budapest; IGPH, Instituto Geologico, La Habana; IGPS, Institute for Geology and Palaeontology,
Sendai; MPMS-R, Museo paleontologico 'Mario Sanchez Roig’, La Habana; NSMT, National Science
Museum, Tokyo; PIMUZ, Palaontologisches Institut und Museum, Universitat Zurich; SGM, State
Geological Museum, University of Sofia; SGNB, Servicio Geologico Nacional, Bogota.

Morphological abbreviations. The following abbreviations are used: D, diameter; H, whorl height; R, ribs per
whorl (of half whorl R/2); U, umbilical diameter; W, whorl width. Suture symbols, applied according to
Wedekind’s (1916) suture terminology (see Kullmann and Wiedmann 1970): E , external lobe; L, lateral lobe;
U, umbilical lobe; /, internal lobe.

Superfamily ancylocerataceae Gill, 1871
Family ancyloceratidae Gill, 1871
Subfamily leptoceratoidinae Thieuloy, 1966

Remarks. Leptoceratoidinae are interpreted to include micromorph heteromorphs with elliptical,
criocone, ancylocone or slightly curved types of coiling, sometimes with a tendency towards torsion,
especially of the initial coil. The suture-line has elements E , L, U , /, poorly incised.

The following genera are included in the Leptoceratoidinae : Veveysiceras gen. nov., Karsteniceras
Royo y Gomez, Hamulinites Paquier, Veleziceras Wright, Eoheteroceras gen. nov., and
Manoloviceras gen. nov.

Most authors refer to Wright (1957) when considering leptoceratids in their entirety to be
members of the Ancyloceratidae. However, Wright split the Colombian genera Karsteniceras and
Veleziceras from Leptoceras , and included the two former genera in the Crioceratitinae. Like
Wright, Manolov (1962) considered Leptoceras to be exclusively Barremian when he proposed
Leptoceratinae as a separate subfamily of Ancyloceratidae for the bulk of leptoceratids. When
Thieuloy (1966) corrected the scope and stratigraphical position of Leptoceras , giving L. brunneri
as the type species and citing a Berriasian age, he proposed the new genus Leptoceratoides , as well

206

PALAEONTOLOGY, VOLUME 37

as Leptoceratoidinae, for the Barremian criocones centred around L. pumilum. Wiedmann (1973)
revived the old genera Karsteniceras and Hamulinites to include all criocone and ancylocone
Barremian micromorphs which were now both included in the Ancyloceratinae. Most successive
authors followed this interpretation, except Immel (1987) who referred Karsteniceras to the
Crioceratitinae, and Gonzales-Arreola and Carrillo-Marti'nez (1987) who referred this genus to the
Helicancylinae Hyatt, 1894.

As a result of the present investigation, the Barremian leptoceratids are considered to represent
a micromorphic stock which separated in the late Hauterivian to early Barremian, presumably from
the main stock of the Crioceratitinae (Text-fig. 7). Three divergent evolutionary lines may have
originated from the basal Lower Barremian Veveysiceras, the ancestor of which is still unknown.
It has to be noted that there is, however, no connection with the Tithonian/Berriasian
Protancyloceratinae Breistroffer.

Distribution. Barremian, mainly Lower Barremian. Southern, eastern and central Europe, Cape Verde Islands,
Central and South America, Japan. The origin of the subfamily (in late Hauterivian time?) remains obscure.
No certain representatives from the Lower Aptian are known.

Genus veveysiceras gen. nov.

Type species. Ancyloceras escheri Ooster, 1860.

Diagnosis. Elliptically coiled, in three arms. After the first planispirally-coiled whorl passing into a
hook, similar to Hamulinites. This is followed by two more elliptical half-whorls. Smooth at first,
the shell on the hooks is covered by fine, dense and simple ribs directed radiate or prorsiradiate on
the lateral side. On the last arm a few weak constrictions are added. Suture-line E , L , U, I with
simplified elements.

Remarks. Coiling is more irregular than in the other genera of the subfamily. Starting with a hook
which is followed by an elliptical whorl, Veveysiceras is characterized by more arms than are present
in Hamulinites and Karsteniceras. Moreover the sculpture is much finer.

Distribution. Lower Barremian of Switzerland, southern France and the Western Carpathians of Slovakia.

Veveysiceras escheri (Ooster, 1860)

Plate 1, figures 1-3; Text-figure 1a

1860 Ancyloceras Escheri Ooster, p. 29 [partim], pi. 37, figs I, 7-9, ?6; non fig. 2 [= Hamulinites
fragilis (Uhlig)] ; non figs 3 4 [= Karsteniceras pumilum (Uhlig)] ; non fig. 5 [ = Crioceratites ? sp.].

EXPLANATION OF PLATE 1

Figs 1-3. Veveysiceras escheri (Ooster). I, NMB 5721, holotype; Hauterivian/Barremian boundary; Veveyse,
Vaud, Switzerland; x 2. 2, GPIT 1719/1; Lower Barremian; Castellane, Haute Provence, France; x L5.
3, SNM Z 21124; Lower Barremian; Zabukovinske quarry near Lictavska Lucka, Central Western
Carpathian, Slovakia; x 1.

Figs 4-5. Karsteniceras ibericum sp. nov. GPIT 1719/2, holotype; Upper Barremian; Barranco de las
Higueras, Sierra Mediana, Subbetic of Alicante, Spain; lateral and ventral views; x 1-5.

Figs 6-8. Karsteniceras beyrichoide sp. nov.; Upper? Barremian; Nydek, Outer Carpathians, Czech Republic.

6, GBA W 3902, x 1 . 7-8, GBA W 3911, holotype ; lateral and oblique view, showing the slight torsion ; x 1 .
Fig. 9. Karsteniceras pumilum (Uhlig). BSM AS III 96, lectotype; Lower Barremian; Straconka, Outer
Carpathians, Poland; x 1.

Arrows indicate position of last suture-line.

PLATE 1

VASICEK and WIEDMANN, Veveysiceras, Karsteniceras

208

PALAEONTOLOGY, VOLUME 37

1902 Crioceras ( Leptoceras ) Escheri Ooster; Sarasin and Schondelmayer, p. 148 | partim], pi. 19, fig.
6; non fig. 4 [= Hamulinites fragilis] ; non fig. 5 [= Crioceratitesl sp.].

Lectotype (designated herein). NMB 5721 (coll. Ooster); Lower(?) Barremian; Veveyse near Chatel St Denis,
Switzerland; here refigured as Plate 1, figure 1, originally figured by Ooster (1860, pi. 37, fig. 1).

Other material. GPIT 1719/14 from the Lower Barremian of Castellane, France; SNM Z-21124 and
unregistered fragments from Krizna Nappe, Slovakia.

Diagnosis. As for genus.

Description. To the generic description can be added that in the holotype the first hook (= Hamulinites)
measures 20 mm in length, while the maximum diameter ranges between 37 and 40 mm; the small diameter of
the ellipsoid is 30 mm; it measures 33 mm in a Carpathian and 19 mm in a French specimen. Maximum whorl
height of the holotype is 8 mm. This means that there is some variation in shell size and in elliptical to
subcircular coiling. The ornamentation consists of fine to very fine and densely spaced radiate ribs on the
phragmocone and more pronounced ribs on the living chamber. The suture-line (Text-fig. 1a) is only partly
visible, L and U are moderately incised and tripartite, the saddles asymmetric bipartite.

Occurrence. The type material is from the Hauterivian/Barremian boundary at Veveyse near Chatel St Denis,
External Prealps, Switzerland. New specimens were collected from the Lower Barremian of Castellane,
Vocontian Trough, France, and the basal Barremian of Lietavska Lucka, Trstie (Krizna Nappe), Central
Western Carpathians, Slovakia.

Genus karsteniceras Royo y Gomez, 1945
[= Leptoceratoides Thieuloy, 1966, p. 289].

Type species. Ancyloceras beyrichi Karsten, 1858.

Revised diagnosis. Small criocones with planispiral to indistinctly trochospiral initial coil, thereafter
loosely coiled. Initial coil smooth, then with simple ribs increasing in strength. Ribs generally
crossing venter; in some species, rib-weakening on the siphonal line leads to the development of a
ventral furrow. The later forms may have marginal tubercles or at least swellings on both sides of
the furrow either on each rib or periodically. Periodic constrictions may be present or not. Suture-
lines with simplified elements E , L, U , /.

Remarks. Leptoceratoides is a synonym of Karsteniceras. In contrast to the opinion of Matsukawa
(1987), we regard the existence of constrictions as insignificant for generic separation in
karsteniceratids.

Karsteniceras differs from most other members of this subfamily in its criocone coiling.
Veveysiceras gen. nov. with its elliptical coiling is closest, but differs in its fine and dense ribbing.
Small-sized or internal whorls of crioceratitids can easily be mistaken for leptoceratids but have
much more complicated lobes and saddles from the very beginning. Moreover, the main ribs of
crioceratitids often carry from one to three tubercles.

The following species are now included in Karsteniceras'. K. asiaticum (Yabe and Shimizu),
K. balernaense Rieber, K. beyrichi (Karsten), K. beyrichoide sp. nov., KP. fUicostatum (Stahlecker),
K. ? heeri (Ooster), K. hoheneggeri sp. nov., K. ibericum sp. nov., K. obatai Matsukawa, K. polieri
Myszyriski, K. pumilium (Uhlig), K. subtile (Uhlig), and K. trinidadense sp. nov.

Species can be separated by differences in mode of ribbing and ventral sculpture. Ooster’s species
‘ Ancyloceras ’ brunneri and ‘ Ancyloceras’’ studeri are true Berriasian to Lower Valanginian
Leptoceras , whereas ‘ Ancyloceras ' escheri Ooster is the type species of Veveysiceras gen. nov.

Uhlig’s species L. assimile , L. fragile and L. parvulum are now transferred to Hamulinites. Of
d’Orbigny’s (1842) small-sized criocones, ‘'Crioceras' puzosianum is unidentifiable. No similar

209

VASICEK AND WIEDMANN: B A R REM IAN AMMONITES

u

L

text-fig. 1 . Leptoceratoid suture-lines, a, frag-
mentary suture-line of Veveysiceras escheri (Ooster)
at H = 2 mm; NMB 5721. b, Karsteniceras beyrichi
(Karsten), external part at H = 5-2 mm. internal lobe
at H = 4 mm; USNM 18609. c, K. ibericum sp. nov.,
complete suture-line of holotype, GPIT 1719/2, at
H = 3-2 mm.

B

U L

E

c

specimen can be located in d’Orbigny’s collection (Sarkar 1955, p. 160, and personal enquiries);
nevertheless, Sarkar (1955) cited this species as characteristic of his genus Spathicrioceras.
‘ Crioceras' cristatum d’Orbigny differs from Karsteniceras in its more complicated suture-line. It
was also cited by Sarkar (1954) as the type species of another new genus, EscragnoUeites , which was
later referred to Imerites Roukhadze by Wright (1957).

The position of 'Ancyloceras' pugnairei Astier, 1851 remains uncertain due to the lack of similar
specimens and its unknown suture-line. 1 To.xoceras' cirtae , described by Coquand (1880) from the
Barremian of Djebel Ouach, was later assigned to Leptoceras (Sayn 1891 ; Kilian 1910). A syntype
from the same locality shows that T. cirtae is similar to Karsteniceras but has a more complicated
suture-line.

Distribution. Karsteniceras is a Barremian genus. It is widespread at the northern margin of the European
Tethys (Text-fig. 9), and ranges into the Caribbean and the Japanese Islands.

Karsteniceras beyrichi (Karsten, 1858)
Plate 2, figures 1-2; Text-figure 1b

1858
non 1945

71954
non 1966

non 1967

Ancyloceras Beyrichii Karsten, p. 103, pi. 1, fig. 4a d.

Karsteniceras beyrichi (Karsten); Royo y Gomez, p. 461, pi. 71, fig. 1 a-c, text-fig. 1
[= Karsteniceras ibericum sp. nov.].

Leptoceras cf. L. beyrichii (Karsten); Imlay, p. 664, pi. 74, figs 23-24.

Leptoceras beyrichi (Karsten); Breskovski, p. 79, pi. 6, fig. 1 [? = Karsteniceras hoheneggeri sp
nov.].

Karsteniceras beyrichi (Karsten); Dimitrova, p. 38, pi. 12, fig. 6 [? = Karsteniceras hoheneggeri
sp. nov.].

210

PALAEONTOLOGY, VOLUME 37

1968 Karsteniceras beyrichi (Karsten); Etayo Serna, p. 54 ( partim ), pi. 1, figs 5, 7; non figs 1-3, text-
fig. 1 [= Karsteniceras ibericwn sp. nov.].
non 1986 Karsteniceras beyrichi (Karsten); Darga and Weidich, pi. 3, fig. 3.

71987 Karsteniceras beyrichi (Karsten); Gonzales-Arreola and Carillo-Martinez, p. 174, fig. 3.
non 1987 Karsteniceras beyrichi (Karsten); Immel, p. 118, pi. 12, fig. 6.

Holotype. The specimen, probably lost, figured by Karsten (1858, pi. 1, fig. 4); Barremian; Velez, Colombia.

Material. One nearly complete specimen, USNM 18609; from Velez-Chipata, Colombia.

Revised diagnosis. Criocone with oval to round whorl section. At the beginning ribs are uniform,
simple and of equal width as intervals. After a diameter of 20 mm, one or two stronger ribs alternate
with bundles of 1-4 less pronounced ribs. Initially all ribs with marginal tubercles, and crossing
venter. Sometimes with a ventral depression. In some specimens, looping of strong ribs between the
tubercles. Suture poorly incised; E with broad median saddle, broad L, narrow and asymmetric U;
saddles asymmetrically bipartite.

Measurements. At D = 22 mm, the following measurements were made on specimen USNM 18603: H = 6 mm
(0-27), W = 6 mm (0 27), U = 8 5 mm (0 39). At Dmax = 24 mm, 35 R/2.

Remarks. From the Colombian specimens previously published (Karsten 1858; Etayo Serna 1968)
and described in this paper, we can conclude that K. beyrichi has a highly variable ribbing. This is
less pronounced in the holotype which is probably lost. The holotype shows a broad and flattened
venter, but no furrow between the indistinct ventrolateral tubercles, as in the specimen figured here
in Plate 2, figure 2. The specimens figured by Royo y Gomez (1945) and some of those of Etayo
Serna (1968) have a distinct ventral furrow and are considered to belong to K. ibericwn sp. nov.

Karsteniceras beyrichioide sp. nov. can only be separated in its adult stage during which the
ribbing remains uniform. The other species of Karsteniceras are easily distinguishable by the
uniformity of ribs, or by the lack of tubercles which exist in some species only for a very short
period.

The specimens referred to K. beyrichi by Imlay (1954) and Gonzales-Arreola and Carrillo-
Martinez (1987) are juveniles or are too fragmentary to be included with certainty in this species.

The specimens figured by Darga and Weidich (1986) and Immel (1987) from the Austroalpine
Lower Barremian are different from K. beyrichi and even the bulk of Karsteniceras species in the
presence of marginal spines where some of the ribs amalgamate. Moreover, their suture-lines are
unknown. Due to these reasons, both specimens are excluded from the karsteniceratids.

Occurrence. Karsteniceras beyrichi is known in the Lower to Upper Barremian transition from Velez-Chipata,
Colombia, and questionably from the Santuario Formation of Maravillas, Mexico, and from Tompire Bay,
Trinidad.

EXPLANATION OF PLATE 2

Figs 1-2. Karsteniceras beyrichi (Karsten). USNM 18609; Lower/Upper Barremian boundary; Velez - Chipata
road, Santander Department, Colombia; lateral and ventral views; x 2.

Figs 3-4. Karsteniceras pumilum (Uhlig). 3, BSM AS III 98; Lower Barremian; Straconka, Outer Carpathians,
Poland; x 2. 4, NMB 5725; Lower Barremian; Veveyse, Vaud, Switzerland; x 1.

Figs 5-8. Karsteniceras subtile (Uhlig). 5, GBAW 3949, lectotype; Lower Barremian; Skalice, Outer
Carpathians, Czech Republic; x 2. 6, GBAW 3901 ; Lower Barremian; Nydek, Outer Carpathians, Czech
Republic; x 2. 7, GBAW 3948; Lower Barremian; Skalice, Outer Carpathians, Czech Republic; x 2.
8, GPIT 1719/3; Lower Barremian; Sierra del Valle, Subbetic of Cadiz, Spain; x 2.

Fig. 9. Karsteniceras hoheneggeri sp. nov. BSM AS III 179, holotype; Upper? Barremian; Malenovice, Outer
Carpathians, Czech Republic; x 1.

PLATE 2

VASICEK and WIEDMANN, Karsteniceras

212

PALAEONTOLOGY, VOLUME 37

text-fig. 2. Karsteniceras ibericum sp. nov. Varia-
bility in E at H between 7 and 7-5 mm; GPIT
1719/11.

Karsteuiceras ibericum sp. nov.

Plate 1, figures 4—5; Text-figures lc, 2

1945 Karsteniceras beyrichi (Karsten); Royo y. Gomez, p. 461, pi. 71, fig. 1; text-fig. 1.

1968 Karsteniceras beyrichi (Karsten); Etayo Serna, p. 54 \partim], pi. 1, figs 1-3; text-figs 4, 8-9; non
text-figs 5, 7 [? = K. beyrichi (Karsten)].

1978 Karsteniceras beyrichi (Karsten); Wiedmann, pi. 4, fig. 2.

Holotype. GPIT 1719/2 (PI. 1, figs 4—5); Upper Barremian; Barranco de las Higueras, Sierra Mediana,
Subbetic, Spain.

Other material. Three paratypes (GPIT 1719/11-13); Barremian; Villa de Leiva, Colombia.

Diagnosis. Criocone with rounded whorl section, simple uniform ribs with marginal tubercles or
thickening of ribs, and distinct siphonal furrow.

Description. The rounded whorl section of the criocone whorl is somewhat broader than high. The simple ribs
are prorsiradiate at first (except the smooth first quarter whorl) changing finally to rectiradiate and rursiradiate.
At first, ribs cross the venter uninterrupted; after one and a half whorls the first marginal tubercles or
thickenings appear. At adult age, ribs are of unequal strength; the ventral furrow on the living chamber
disappears. The last whorls have 42-50 ribs. The suture-line (Text-fig. lc) is usually poorly incised and L is
broadly rounded. E has a distinct median saddle which can, however, vary in width (Text-fig. 2). The other
lobes are undivided. The saddles are symmetrically bipartite.

Measurements. The holotype measures 17 mm in maximum diameter (last suture at D = 10-5 mm). The largest
specimen known (Etayo Serna 1968, pi. 1, fig. 1) has a final D of 40 mm.

Remarks. Karsteniceras ibericum sp. nov. is easily distinguished from its relatives by the distinct
ventral furrow and the density of ribs.

213

VASICEK AND WIEDMANN: BARREMIAN AMMONITES

Occurrence. Known from the Upper Barremian of Sierra Mediana, Subbetic, southern Spain, from the Lower
to Upper Barremian transition at Ranzenberg, Drusberg Beds, Vorarlberg, Austria, and from deposits of a
similar age at Villa de Leiva, Colombia.

Karsteniceras beyrichoide sp. nov.

Plate 1, figures 6-8, Text-figure 3a

1883 Crioceras ( Leptoceras ) Beyrichi Karst.; Uhlig, p. 272, pi. 32, figs 4-6, 78.

1960 Leptoceras beyrichi Karsten; Drushtchic, p. 295, pi. 40, fig. 4.

1976 Karsteniceras aff. beyrichi (Karsten); Avram, p. 34, pi. 3, fig. 9.

Holotype. GBAW 391 1 (PI. 1, fig. 7); Upper(?) Barremian; Nydek, Silesian Unit, Czech Republic. The original
of Crioceras ( Leptoceras ) beyrichi (sensu Uhlig 1883, pi. 32, fig. 4).

Other material. GBAW 3902, 1883/4/1 18, BSM AS III 178, the three specimens figured by Uhlig (1883, pi. 32,
figs 5-6, 8).

Diagnosis. Small-sized criocones with weak torsion of the shell. Whorl section subrectangular,
probably as broad as high. The lateral sides are flat, as is the broad venter which carries a shallow
siphonal furrow. Initial coil unknown; thereafter, sculpture consists of single sharp ribs as broad
as the intervals. On the living chamber ribbing becomes less crowded. Marginal tubercles may be
present or not; they are, moreover, of variable strength. Suture-line (Text-fig. 3a) is very simple with
goniatitic lobes and bipartite saddles.

Measurements. The holotype has a final diameter of 39 mm, half whorl (?) of living chamber included,
Hmax = 10 mm. The last whorl is covered by 61 ribs. The most complete specimen, GBAW 3902 (Uhlig 1883,
pi. 39, fig. 5), has a Dmax = 30 mm; at D 26 3 mm, H is equivalent to 6-4 mm, U = 16 mm. The last whorl
carries 64 ribs.

Remarks. Restudy of Uhlig’s (1883, pi. 32, fig. 5) specimen GBAW 3902 has shown that it is formed
by two living chambers placed in opposite directions, both of the same species. One of these
specimens shows parts of the last suture-line. Another specimen (GBAW 1883/4/118; Uhlig 1883,
pi. 32, fig. 8) cannot be included in K. beyrichoide with certainty; the ribs are less frequent than usual
and carry pronounced marginal tubercles.

The present species can be distinguished from K. beyrichi by the differing ribbing on the living
chamber and near the mouth border.

Occurrence. At present, K. beyrichoide sp. nov. is known with certainty only from the Upper(?) Barremian of
Nydek, Silesian Unit, Outer Carpathians, Czech Republic. Doubtfully included are specimens, possibly from
the same stratigraphical level, from the Outer Dacidian Nappe, Roumania, and from the Kuma River, North
Caucasus.

Karsteniceras pumilum (Uhlig, 1883)

Plate 1, figure 9; Plate 2, figures 3-4

1860 Ancyloceras Esclieri Ooster, pi. 37, figs 3^1 only.

1883 Crioceras ( Leptoceras ) pumilum Uhlig, p. 270, pi. 29, figs 4-6.

1902 Crioceras ( Leptoceras ) pumilum Uhlig ; Sarasin and Schondelmayer, p. 147, pi. 20, fig. 4.

71926 Leptoceras cfr. pumilum Uhlig; Yabe et a /., p. 73, pi. 15, fig. 20.

1938 Leptoceras parvulum Uhlig; Roman, pi. 35, figs 335-336.

1957 Leptoceras pumilum Uhlig; Wright, p. L211.

1958 Leptoceras parvulum Uhlig; Luppov and Drushtchic, pi. 48, figs 6-7.

1962 Leptoceras pumilum Uhlig; Manolov, p. 532.

71962 Leptoceras pumilum Uhlig; Akopyan, p. 205, pi. 1, fig. 5.

1966 Leptoceratoides pumilus (Uhlig); Thieuloy, p. 289.

214

PALAEONTOLOGY, VOLUME 37

1969 Leptoceras pumilus Uhlig; Wiedmann, pi. 2, fig. 2.

1972 Leptoceras pumilus (Uhlig); Vasicek, p. 54, pi. 4, fig. 5.

1972 Leptoceras pumilum Uhlig; Wiedmann, pi. 1, fig. 2.

1976 Leptoceratoides pumilus (Uhlig); Avram, p. 33, pi. 4, fig. I.

1984 Leptoceras pumilum Uhlig; Avram and Kusko, p. 14, pi. 2, fig. 8.

1987 Leptoceratoides pumilus (Uhlig); Matsukawa, p. 349.

1990 Karsteniceras subtile (Uhlig); Vasicek, pi. 1, fig. 6.

Lectotype (designated by Thieuloy 1966). BSM AS III 96; Lower Barremian; Straconka, Poland. The original
of Crioceras ( Leptoceras ) pumilum Uhlig (1883, pi. 29, fig. 4; PI. 1, fig. 9).

Other material. The two paralectotypes of Uhlig (1883), specimens figured by Ooster (1860), and new
discoveries from the Outer Carpathians (OSM B 13035, VSB T 4/17, and unregistered specimens).

Revised diagnosis. Becoming criocone after an advolute inner whorl. For a short period (at D =
12-15 mm) whorls may touch each other. First whorl smooth, becoming ribbed at 8 mm D. First
shallow constrictions appear at D = 10-12 mm. While simple ribs are radiate at first, they change
later to rursiradiate. Exceptionally some ribs bifurcate on the lateral side of living chamber.
Constrictions increase in strength. Suture-line unknown.

Measurements. The specimens attain diameters of 30-35 mm.

Remarks. Karsteniceras pumilum differs from all other species of the genus in its frequent and
distinct constrictions. K. subtile (Uhlig) is similar but has less frequent and less pronounced
constrictions; ribbing starts earlier (D = 4 mm); looping of ribs occurs only in some specimens; also
the final size is smaller.

The Japanese specimen (Yabe et al. 1926) shows marginal tubercles in addition to the specific
characters of this species; its inclusion remains, therefore, uncertain.

Occurrence. Ticha and Trojanovice, Outer Carpathians of the Czech Republic; Straconka and Gorki Wielkie,
Outer Carpathians of Poland; Getic and Outer Dacididn Nappes, Carpathians of Roumania; and Veveyse,
External Prealps of Switzerland. Uncertain are the occurrences in America and Japan. The age is undefined
Barremian; the Silesian specimens seem to belong only to the lower portion of the Barremian.

Karsteniceras subtile (Uhlig, 1883)

Plate 2, figures 5-8; Text-figure 3 b

1883 Crioceras ( Leptoceras ) subtile Uhlig, p. 271, pi. 29, figs 7-9.

1966 Leptoceratoides subtilis (Uhlig); Thieuloy, p. 289.

? 1967 Leptoceras subtile Uhlig; Dimitrova, p. 39, pi. 12, fig. 8; non fig. 7 [= Karsteniceras balernaense
Rieber]; text-fig. 18.

1972 Leptoceratoides subtilis (Uhlig); Vasicek, p. 54, pi. 7, fig. 4.

1972 Leptoceratoides cf. subtilis (Uhlig); Vasicek, pi. 5, fig. 4; text-fig. 16.

1976 Leptoceratoides subtilis (Uhlig); Avram, p. 33.

71977 1 Karsteniceras cf. subtilis (Uhlig); Myczyriski, p. 155, pi. 4, fig. 5.

1984 Leptoceras subtile Uhlig; Avram and Kusko, p. 14, pi. 2, figs 6-7.

non 1990 Karsteniceras subtile (Uhlig); Vasicek, pi. 1, fig. 6 [ = Karsteniceras pumilum (Uhlig)].

1991 Karsteniceras subtile (Uhlig); Manthey, pi. 2, fig. 1.

Lectotype (designated by Dimitrova 1967). GBAW 3949; Lower Barremian; Skalice, Czech Republic. The
original of Crioceras ( Leptoceras ) subtile Uhlig (1883, pi. 29, fig. 9; PI. 2, fig. 5 herein).

Other material. The two paralectotypes (GBAW 3948, 3901) of Uhlig (1883), and new discoveries from the
Czech Carpathians (VSB T 5/172), Eastern Alps (GPIT 1719/15) and the Subbetic of southern Spain (GPIT
1719/3).

215

VASICEK AND WIEDMANN: BARREM IAN AMMONITES

text-fig. 3. Leptoceratoid suture-lines, a, Karst-
eniceras beyrichoide sp. nov., fragmentary suture-line
at H = 6 mm; BSM AS III 178. b, K. subtile (Uhlig),
external suture-line of lectotype at H = 2-2 mm;
GBAW 3949. c, K. hoheneggeri sp. nov., fragmentary
suture-line at H = 3mm; BSM AS III 179. d,
K. hoheneggeri, external suture-line at H = 6 mm ;

BSM AS III 454.

u

E

Revised diagnosis. Small criocones with rounded whorl section. Ribbing starts at D = 4 mm and
consists of simple and looped ribs; at later stages ribs show ventrolateral thickening and seem to
weaken on the venter. Near the mouth border ribbing becomes fine and dense. Weak constrictions,
especially in early ages. Suture-line with trifid L and U and bipartite saddles (Text-fig. 3b).

Measurements. The lectotype has a maximum D of 19 mm. At D = 15, Hmax = 4 mm (0-27), U = 8-5 (0-57).
24 R/2; in some specimens even more ribs occur.

Remarks. Special characters of K. subtile are its very small size, the occasional looping of ribs, and
the dense and fine ribbing near the mouth border.

Occurrence. Karsteniceras subtile is presumably restricted to the Lower Barremian. Regionally, it is known
from Ticha, Skalice and Nydek, Outer Carpathians, Czech Republic; probably from Lietavska Lucka, Central
Carpathians of Slovakia; from the Getic and Outer Dacidian Nappes, Carpathians of Roumania; from Bad
Ischl in the Austroalpine of Salzkammergut, Austria; and from Sierra del Valle, Subbetic of Spain. Published
records from Bulgaria and Cuba are uncertain.

Karsteniceras hoheneggeri sp. nov.

Plate 2, figure 9; Plate 3, figures 1-3; Text-figure 3c-d

1883 Crioceras (Leptoceras) n. sp. ind. aff. cristatum d’Orbigny; Uhlig, p. 272, pi. 32, fig. 3.
1883 Crioceras (Leptoceras) n. sp. ind.; Uhlig, p. 272, pi. 32, fig. 7.

71966 Leptoceras beyrichi (Karsten); Breskovski, p. 79, pi. 6, fig. 1.

71967 Karsteniceras beyrichi (Karsten); Dimitrova, p. 9, pi. 12, fig. 6.

216

PALAEONTOLOGY, VOLUME 37

Holotype. BSM AS III 179; Upper(?) Barremian; Malenovice, Czech Republic. Crioceras (Leptoceras) n. sp.
ind. aff. cristatum (d’Orbigny) of Uhlig (1883, pi. 32, fig. 3; PI. 2, fig. 9).

Other material. BSM AS III 454, Hohenegger Collection, figured by Uhlig (1883). GPIT 1719/4, fragmentary
specimen from Breggia, Southern Alps.

Diagnosis. Regular criocone coiling. Whorl section rounded with inflated lateral sides and flat to
concave venter. In the adult, simple and strong ribs of unequal strength, with marginal thickening
and concave venter. Three constrictions on the last half whorl. Suture-line with simple elements and
broad L.

Measurements. The holotype has Dmax = 38 mm; at D = 28 5 mm, H = 9 mm and U = 13-6 mm. On the last
whorl, thirty eight ribs can be counted.

Remarks. Karsteniceras hoheneggeri sp. nov. is similar to K. beyrichioide sp. nov. and to K. pumilum.
Its ribbing is, however, more pronounced and less crowded. K. trinidadense sp. nov. has a much
finer ribbing with larger interspaces.

Occurrence. Known from the Upper(?) Barremian of Malenovice and Nydek, Silesian Unit, Outer Carpathians,
Czech Republic; and from the Upper Barremian of Breggia, Southern Alps, Switzerland. Uncertain is the
inclusion of specimens from the Upper Barremian of Bulgaria.

Karsteniceras trinidadense sp. nov.

1954 Leptoceras sp. indet.; Imlay, p. 664, pi. 74, figs 16-17; non fig. 18.

Holotype. USMM 108726; Barremian ; Tompire Bay, Trinidad; figured as Leptoceras sp. indet. by Imlay (1954,
pi. 74, fig. 16).

Diagnosis. Small criocones with fine and dense ribbing and with marginal thickening. Ribs widely
spaced, intervals four times as wide as ribs. Suture-line unknown.

Measurements. The holotype has a maximum diameter of 27 mm.

Remarks. The type material is rather poorly preserved in a phyllitic matrix, and is deformed. Whorl
section and suture-line are unknown. The ribbing, which consists of fine and strong ribs with large
interspaces, is however so different from all other Karsteniceras species that a specific separation can
be proposed. The marginal thickening of ribs is not as obvious as mentioned by Imlay (1954). This
is a typical Karsteniceras , even if the suture-line is unknown.

Occurrence. Lower part of Upper Barremian of Tompire Bay, Trinidad.

EXPLANATION OF PLATE 3

Figs 1-3. Karsteniceras hoheneggeri sp. nov. 1-2, BSM AS III 454; Upper? Barremian; Nydek - Ostra hora.
Outer Carpathians, Czech Republic. 1, lateral and ventral views; x 1. 3, GPIT 1719/4; Upper Barremian;
Breggia, Ticino, Switzerland; x 3.

Fig. 4. Karsteniceras balernaense Rieber, GPIT 1719/5; Upper Barremian; Breggia, Ticino, Switzerland; x 1.
Figs 5-12. Hamulinites parvulus (Uhlig). 5, BSM AS III 453. lectotype; Lower Barremian; Verovice, Outer
Carpathians, Czech Republic; x 1. 6, GPIT 1719/6; Lower Barremian; Bad Ischl, Salzkammergut, Austria;
x3. 7. VSB Pi 1/12; Lower Barremian; Pindula near Frenstat p. R., Outer Carpathians, Czech Republic;
x 2. 8, OSM B 13664; Lower Barremian; Angles, Haute Provence, France; x 2. 9, MHNP, d’Orbigny coll.
5428-1; Barremian; Escragnolles, Var, France; x L5. 10, SNM Z 21 125; Lower Barremian; Lietavska
Lucka, Central Western Carpathians, Slovakia; x 1. 11, GPIT 1719/7; Lower Barremian; La Querola near
Cocentaina, Prebetic of Alicante, Spain; x 2. 12, GPIT 1719/8, same horizon and locality as fig. 11; x 3.

PLATE 3

VASICEK and WI EDM ANN, Karsteniceras, Hamulinites

PALAEONTOLOGY, VOLUME 37

Karsteniceras balernaense Rieber, 1977

Plate 3, figure 4

1967 Leptoceras subtile Uhlig; Dimitrova, p. 39 [partim\, pi. 12, fig. 7; non fig. 8 [? = K. subtile Uhlig].
1977 Karsteniceras balernaense Rieber, p. 779, pi. 1, figs 1-7, text-fig. 2.

71984 Leptocerasl cf. barnaense (Rieber); Avram and Kusko, p. 14, pi. 2, fig. 9.

Holotype. PIMUZ L/1584; Barremian; Balerna, Switzerland; figured by Rieber (1977, pi. 7, fig. 1).

Other material. GPIT 1719/5; a poorly preserved specimen from the southern Alps of Breggia Gorge,
Switzerland.

Revised diagnosis. Small criocones, probably with weak torsion. Single sharp uniform ribs from a
D of 3-5 mm. 52 ribs per whorl. No tubercles. Venter unknown. Suture-line simple.

Remarks. The uniform type of ribbing up to the mouth border separates K. balernaense from
K. subtile ; the lack of constrictions distinguishes this species from K. pumilum. K. polieri Myczyriski
(1977) differs in having bifurcating ribs.

Occurrence. Originally described from the Upper Barremian of Breggia Gorge near Balerna, Ticino,
Switzerland; known also from the Lower Barremian of Kraptshene, Prebalkan, Bulgaria, and probably the
Romanian Southern Carpathians.

Karsteniceras polieri Myczyhski, 1977

1977 Karsteniceras polieri Myczyhski, p. 154, pi. 4, figs 1, 3, 7.

Holotype. IGPH No. 2556b; Lower Barremian; Polier, Cuba; figured by Myczyriski (1977, pi. 4, fig. 7).

Revised diagnosis. Small criocones with dense and fine ribbing. The prorsiradiate ribs bifurcate
periodically at the umbilical border. No constrictions.

Remarks. This species resembles both K. subtile and K. balernaense but it differs in having
bifurcating ribs.

Occurrence. Lower Barremian of the Polier Formation, Sierra del Rosario, Cuba.

Karsteniceras asiaticum (Yabe and Shimizu, 1926)

1926 Leptoceras asiaticum Yabe and Shimizu in Yabe et al ., p. 73, pi. 15, fig. 21.

1966 Leptoceratoides asiaticus (Yabe and Shimizu); Thieuloy, p. 289.

1988 Karsteniceras asiaticum (Yabe and Shimizu); Matsukawa, p. 399, fig. 3: 4-6; fig. 4.

Holotype. IGPS 22849; Lower Barremian; Ishido, Japan; figured by Yabe and Shimizu (in Yabe et al. 1926,
pi. 15, fig. 21).

Revised diagnosis. Small criocones with subrectangular whorl section. Ribs are fine and dense at the
beginning, later they show broad interspaces and become rursiradiate on the living chamber. At
later stages, about twenty ribs per half whorl. Ventral side with weak depression.

Remarks. While ribs cross the ventral depression in K. asiaticum , they are interrupted on the venter
of the similar K. obatai Matsukawa.

Occurrence. Lower Barremian of Ishido, Hondo, Japan.

219

VASICEK AND WIEDMANN: BARREMI AN AMMONITES

Kcirsteniceras ohatai Matsukawa, 1987

1926 Ancyloceras! sp. indet.; Yabe el al., p. 71, pi. 15, figs 12-13.

1987 Kcirsteniceras obatai Matsukawa, p. 349, figs 1-2; 3, 1-4; 4.

1988 Karsteniceras obatai Matsukawa; Matsukawa, p. 401, fig. 3: 7-9; fig. 5.

Holotype. NSMT-PM 9589; Lower Barremian; Isejigauara, Japan; figured by Matsukawa (1987, fig. 3: 3).

Diagnosis. Small criocones with smooth initial whorl. From a diameter of 10 mm simple and
rectiradiate ribs with marginal tubercles. Ribs are interrupted at a narrow siphuncular furrow.
Suture elements simplified.

Remarks. In Karsteniceras obatai marginal tubercles and the siphuncular furrow disappear on the
living chamber. The differences between K. obatei and K. asiaticwn are given above.

Occurrence. Lower Barremian of Isejigauara, Hondo, Japan.

Karstenicerdsl filicostatum (Stahlecker, 1935)

1935 ‘ Toxoceras' filicostatum Stahlecker, p. 286, pi. 14, fig. 16.

1966 Leptoceratoides filicostatus (Stahlecker); Thieuloy, p. 289.

Holotype. GPIT 493/14/16; Lower(?) Barremian; Maio, Cape Verde Islands; figured by Stahlecker (1935,
pi. 14, fig. 16).

Revised diagnosis. Middle-sized criocones with open coil. Sculpture consists of fine, dense and
radiate ribs. Each third/fourth rib with weak marginal tubercles. Suture-line unknown.

Remarks. Karsteniceras! filicostatum resembles the likewise dubious K. ? heeri (Ooster) in coiling.
However, the latter has coarser ribbing. Like K .? heeri, the present species is too poorly known (the
suture-line is unknown) to be included in Karsteniceras with certainty.

Occurrence. Lower(?) Barremian of Maio, Cape Verde Islands.

Karsteniceras ? heeri (Ooster, 1860)

1860 Ancyloceras Heeri Ooster, p. 32 [par tint], pi. 38, figs 5, ? 1-3; non fig. 4 [= Anahamulina distans
Vasicek].

1902 Crioceras ( Leptoceras ) Heeri Ooster; Sarasin and Schondelmayer, p. 149 \partim], pi. 20, fig. 2;
non fig. 3 [ = A. distans Vasicek],

Lectotype (designated herein). NMB 5683; Barremian; Veveyse, Switzerland; figured by Ooster (1860, pi. 38,
fig. 5).

Revised diagnosis. Middle-sized criocones(?), whorl slowly increasing in height. Ribs as broad as
interspaces, radiate on lateral sides, becoming prorsiradiate and weakly inflated at the umbilical
margin. Last whorl with weak constrictions. Suture-line unknown.

Measurements. Dmax = 40 mm, H = 8 0 mm (0 20), U = 27 5 mm (0 69). R/2 > 50.

Remarks. Due to the fragmentary preservation of the paralectotypes, it is unknown whether
K.l heeri is criocone throughout or has the elliptical coiling of Veveysiceras. In this case, Ooster’s
(1860, pi. 38, figs 1-3) juveniles would have to be included.

According to Sarasin and Schondelmayer (1902, p. 149), some specimens of Ancyloceras
sabaudianum Pictet and Loriol (1858, pi. 6, figs 6, 9) should be referred to K.l heeri ; their juvenile
coils are, however, criocone.

220

PALAEONTOLOGY, VOLUME 37

K.l filicostatus (Stahlecker) has a coiling similar to the present species, but has a finer and denser
ribbing.

Occurrence. Barremian of Veveyse, External Prealps, Switzerland.

Genus hamulinites Paquier, 1901

[= Leptoceras Uhlig, 1883 [partim] ; Eoleptoceras Manolov, 1962; E. ( Tzankoviceras ) Manolov, 1962;

E. ( Wrightites ) Manolov, 1962.]

Type species. Hamulina munieri Nickles, 1894 [ = Leptoceras parvulum Uhlig, 1883.].

Revised diagnosis. Small-sized ancylocones; first two whorls advolute, then slowly uncoiling
(proversum according to Vasicek 1972, fig. 12) towards the final hook which is mainly formed by
the living chamber. Simple ribs without tuberculation. Suture-line with four elements ( E , L, C7, /),
with relatively simple outlines.

Remarks. Hamulinites comprises small-sized ancyloceratids, but there are no early criocone whorls
and no marginal spines; suture elements remain simple, but are more complicated than in
Karsteniceras. Due to its ancylocone coiling, Hamulinites is easy to separate from the other
representants of the Leptoceratoidinae. There is, however, a certain similarity to small-sized
hamulinids, above all to Hamulina varusensis d'Orbigny, 1850 (see d’Orbigny 1852, p. 221, pi. 5, figs
4-6; Cottreau 1937, p. 72, pi. 80, figs 3-6). After studying the four syntypes of H. varusensis
preserved in the MHNP, we can assume that d'Orbigny’s complete specimen (1852, pi. 5, fig. 4) is
a reconstruction based on the fragments known. These bear a complicated suture-line (see Cottreau
1937, p. 72) of hamulinid configuration. H. varusensis is, therefore, referred to Anahamulina Hyatt.
The most complete syntype (Cottreau, 1937, pi. 80, fig. 4), however, is different from d’Orbigny’s
figures and is here included in Hamulinites parvulus Uhlig.

Anahamulina distans Vasicek (1972, pi. 15, fig. 4) also shows similarity with Hamulinites in size
and coiling, but has a very different, more complicated suture-line (see Text-fig. 4). This complicated

L E text-fig. 4. Fragmentary suture-line of Anahamulina

distans Vasicek, at H = 6-5 mm; BSM AS III 452.

suture-line, generally larger size and, in most species, constrictions on the hook, are the main
differences distinguishing Anahamulina from Hamulinites.

The middle-sized Ancyloceras sabaudianum Pictet and Loriol (1858) also exhibits similarities with
the genus Hamulinites , as does ‘ Hamulinites ’ norteyi Myczyriski and Triff, 1986. However, the
former shows a criocone initial coil and marginal spines unknown from Hamulinites , while the latter
exhibits a slightly heteroceratid coiling and is therefore included in Eoheteroceras gen. nov.

Another doubtful group of forms was described by Drushtchic (1960, pi. 39, fig. 3) as ‘ Leptoceras
biplex' von Koenen. These are small-sized ancylocones with true ancyloceratid sculpture, i.e. three
rows of tubercles on the ribs of both arms. The position of these forms cannot be clarified without

221

VASICEK AND WIEDMANN: BARREMI AN AMMONITES

studying Russian specimens. As a consequence, only the following species remain attributable to
Hamulinites : H. parvulus (Uhlig), H. fragilis (Uhlig) and H. assimilis (Uhlig).

Distribution. Mainly Lower Barremian, rare in the Upper Barremian; Outer Western Carpathians of the Czech
Republic and Poland; Central Carpathians of Slovakia; Carpathians of Roumania; Bulgaria; Austroalpine of
Austria; Prealps of Switzerland; Vocontian Trough, France; Subbetic of Spain; Cuba (?).

71860
1883
1894
1902
1937
non 1938
1957
non 1958

1962

1962

1962

1962

1962

1963
1966

1966

1967
1967
1967
1967
1967

non 1967
1972
1976
71986
1990

Hamulinites parvulus (Uhlig, 1883)

PI. 3, figures 5-12; Text-figure 5a

Ancyloceras Fourneti Astier; Ooster, p. 22, pi. 34, fig. 10; non figs 9, 1 1 .

Crioceras ( Leptoceras ) parvulum Uhlig, p. 273, pi. 29, figs 3, 10.

Hamulina Munieri Nickles, p. 59, pi. 5. figs 7-8, text-fig. 42.

Hamulina parvula Sarasin and Schondelmayer, p. 166, pi. 23, figs 4-5.

Hamulina Varusensis d’Orbigny; Cottreau, p. 72, pi. 80, fig. 3; non figs 4-6.

Leptoceras parvulum Uhlig; Roman, pi. 35, figs 335-336 [= Karsteniceras pumilum (Uhlig)].
Hamulinites munieri (Nickles); Wright, p. L215.

Leptoceras parvulum Uhlig; Luppov and Drushtchic, pi. 48, figs 6-7 [= Karsteniceras pumilum
(Uhlig)].

Eoleptoceras ( Wrightites ) parvulum (Uhlig); Manolov, pp. 532, 534, pi. 75, figs 3, 11-12.

7 Eoleptoceras ( Tzankoviceras ) assimilis (Uhlig); Manolov, pi. 74, figs 3M.

Eoleptoceras (Tzankoviceras) tzankovi Manolov, p. 533, pi. 75, figs 2, 7-8.

Eoleptoceras ( Wrightites ) parvulum kraptshenensis Manolov, p. 535, pi. 75, figs 4—6, text-fig. lc.
Eoleptoceras (Wrightites) wrighti Manolov, p. 535, pi. 75, figs 9-10, text-fig. 1b.

Hamulinites munieri (Nickles); Wiedmann, p. 108, pi. 1, fig. 3, text-fig. 2.

Eoleptoceras parvulum (Uhlig); Thieuloy, p. 289.

Anahamulina varusensis (d’Orbigny); Breskovski, p. 82, pi. 4, fig. 2.

Eoleptoceras (E.) parvulum (Uhlig); Dimitrova, p. 36, pi. 17, fig. 7, text-fig. 16.

Eoleptoceras (E.) varusensis (d’Orbigny); Dimitrova, p. 36, pi. 17, fig. 8.

Eoleptoceras (s.lato) wrighti Manolov; Dimitrova, p. 36, pi 16, fig. 8.

Eoleptoceras (Tzankoviceras) tzankovi Manolov; Dimitrova, p. 37, pi. 18, figs 7-9.
Eoleptoceras (Tzankoviceras) assimilis (Uhlig); Dimitrova, p. 37, pi. 18, fig. 10.

Eoleptoceras (Wrightites) parvulum (Uhlig); Nagy, p. 68, pi. 3,- fig. 3 [= Crioceratitmae juv.]
Hamulinites parvulus (Uhlig); Vasicek, p. 53, pi. 7, fig. 2, text-fig. 15.

Hamulinites cf parvulus (Uhlig); Avram, p. 34, pi. 4, fig. 2.

Hamulinites aff. parvulus (Uhlig); Myczynski and Triff, p. 126, pi. 2, figs 1, 9, 15.

Hamulinites parvulus (Uhlig); Vasicek, pi. 1, fig. 4.

Lectotype (designated by Manolov 1962). BSM AS III 453; Lower Barremian; Verovice (Wernsdorf), Czech
Republic; figured by Uhlig (1883, pi. 29, fig. 3).

Other material. OSM B 13 033, VSB Pi 1/12, K 8/014, OT 5/18 and unregistered fragments, about twenty
specimens from the Czech Outer Carpathians; SNM Z-21 125 and unregistered fragments. Central
Carpathians of Slovakia; OSM B 13 664, Angles and MHNP - d’Orbigny coll. 5428-1, Escragnolles, southern
France; GPIT 1719/7, 1719/8, Prebetic of Alicante; GPIT 1719/6, Bad Ischl, Austroalpine and the material
from Sarasin and Schondelmayer (1902) described as 1 Hamulina parvula'.

Revised diagnosis. Small ancylocones, eventually with heteroceratid torsion of the initial coil. Shell
initially smooth, then with single sharp untubercled ribs crossing venter without interruption.
Suture-line with relatively high and smooth median saddle in E\ L and U trifid, saddles simple and
bipartite (Text-fig. 5a).

Remarks. Hamulinites parvulus is a variable species, exhibiting variations in size (D around 30 mm),
coiling of the initial coil, separation of the two final arms, and strength and density of ribbing.
Therefore, a number of Manolov’s (1962) species have to be synonymized with H. parvulus ; extreme

222

PALAEONTOLOGY, VOLUME 37

forms, which have criocone initial coils, were described as ‘ Eoleptoceras assimile' (Uhlig) by
Manolov (1962, pi. 74, figs 3-4). Unfortunately, none of the specimens with unknown suture-lines
can be determined generically.

Hamulina parvula , described by Sarasin and Schondelmayer (1902) from the Barremian of Chatel
St Denis, is very similar in coiling, size and sculpture, but was separated due to the existence of two
rows of tubercles on the living chamber. After revision of the two specimens figured, it is obvious
that there is not even a single row of tubercles. One of the specimens (pi. 23, fig. 4) was, however,
an injured and restored living chamber margin. Moreover, fragments of the suture-line visible
exhibit simplified lobes and saddles. In consequence, these specimens are included in H. parvulus.

H. fragile can easily be separated by its typical reduction of ribbing near the mouth border, while
H. assiniilis has stronger ribs on the larger and more robust shell.

Occurrence. A cosmopolitan and mainly Lower Barremian species which was first described from Verovice and
Lipnik, Outer Western Carpathians of the Czech Republic and Poland. Also recorded from: Escragnolles,
southern France; Querola, Prebetic of southern Spain; Chatel St Denis/Switzerland; Kraptshene,
Prebalkan, Bulgaria; Ceahlau Nappe East Carpathians of Roumania; and Cuba(?). Additionally, we have
collected the species in the lower part of the Barremian type section at Angles, southern France; the Lower
Barremian of Querola, Prebetic of Spain; the Austroalpine beds near Bad Ischl, Austria,; and in the Outer
Carpathians (Ticha, Trojanovice, Pindula), Czech Republic and Central Western Carpathians (Lietavska
Lucka, Butkov, Zrazy), Slovakia.

Hamulinites fragilis (Uhlig, 1883)

Plate 4; figures 1-2; Text-figure 5b

1860 Ancyloceras Escheri Ooster, p. 29 [partim\, pi. 37, fig. 2; non figs 1, 7-9, 76 [= Veveysiceras
escheri (Ooster)]; non figs 3-4 [= Karsteniceras pumilum (Uhlig)]; non fig. 5 [= Crioceratitesl
sp.].

1883 Crioceras {Leptocer as) fragile Uhlig, p. 274, pi. 29, fig. 11.

1902 Crioceras (Leptoceras) Escheri Ooster; Sarasin and Schondelmayer, p. 148 \partim], pi. 19, fig.

4; non fig. 5 [ = Crioceratites ? sp.]; non fig. 6 [= Veveysiceras escheri (Ooster)].

1984 Eoleptoceras (E.) aff. fragile (Uhlig); Avram and Kusko, p. 14, pi. 2, fig. 5.

Neotype (designated herein). OSM-B 13663; Lower Barremian; Nydek, Silesian Unit, Czech Republic
(PI. 4, fig. 1). Uhlig’s type specimen from the same area (Lipnik) can be located neither at GBAW nor at BSM,
and is thus considered lost.

Other material. The NMB specimens which Ooster (1860) described as Ancyloceras escheri.

EXPLANATION OF PLATE 4

Figs 1-2. Hamulinites fragilis (Uhlig). 1, OSM B 13663, neotype; Lower Barremian; Nydek, Outer
Carpathians, Czech Republic; x 2. 2, NMB 5725a; Lower Barremian; Veveyse, Vaud, Switzerland; part
(left) and counterpart (right); x 1.

Fig. 3. Hamulinites assimilis (Uhlig). BSM AS III 177, holotype; Upper Barremian; Mistrovice, Outer
Carpathians, Czech Republic; x 1.

Figs 4-5. Eoheteroceras silesiacum gen. et sp. nov. 4, GBAW 3938, holotype ; Lower Barremian ; Gorki Wielkie,
Outer Carpathians, Poland; x L5. 5, GPIT 1719/9; Lower Barremian, La Querola near Cocentaina,
Prebetic of Alicante, Spain ; x 2.

Figs 6-7. Eoheteroceras uhligi (Vasicek). 6, VSB OT 5/30; Lower Barremian; Ostravice, Outer Carpathians,
Czech Republic; x 2. 7, GPIT 1719/10; Upper Barremian; Breggia, Ticino, Switzerland; x2-5.

Fig. 8. Manoloviceras saharievae (Manolov). OSM B 13039; Lower Barremian; Shaft Frenstat 5, at 294 m
depth. Outer Carpathians, Czech Republic; ‘nest’ of mostly juvenile individuals; x 2.

PLATE 4

VASICEK and WIEDMANN, Hamulinites, Eoheteroceras , Manoloviceras

224

PALAEONTOLOGY, VOLUME 37

E

A

k L U

text-fig. 5. Leptoceratoid suture-lines, a, suture-lines
of Hamulinites parvulus (Uhlig) at H = 3-3 and
3-5 mm; BSM AS III 453. b, fragmentary suture-line
of H. fragilis (Uhlig) at H = 3-5 mm; NMB 5725a.
c, external suture-line of H. assimilis (Uhlig) at
H = 6-2 mm; BSM AS III 177.

Revised diagnosis. Small ancylocones. Initial coil gently curved; with final hook. Shell initially
smooth, then with sharp and fine simple ribs, becoming replaced by fine and dense growth-lines on
the central hook ( = living chamber?). Suture-line of Hamulinites type.

Measurements. The neotype has Dmax = 22 mm, the retroversum measures 15 mm, Hmax = 7 mm. The diameter
of the coiled first whorl is 2 mm.

Remarks. Hamulinites fragilis is easily separated, even when juvenile or fragmentary, from the
similar Veveysiceras escheri (Ooster) by its fine ribbing.

Occurrence. Lower Barremian. Lipnik and Nydek, Outer Western Carpathians, Poland and Czech Republic;
Veveyse, External Prealpes, Switzerland; and probably from Romania.

Hamulinites assimilis (Uhlig, 1883)

Plate 4, figure 3, Text-figure 5c

1883 Crioceras (Leptoceras) assimile Uhlig, p. 272, pi. 32, fig. 9.

1962 Eoleptoceras ( Tzankoviceras ) assimilis (Uhlig); Manolov, p. 533 [partim], pi. 73, figs 7-8; non pi.
74, figs 3 4 [? = Hamulinites parvulus (Uhlig)].

non 1967 Eoleptoceras ( Tzankoviceras ) assimilis (Uhlig); Dimitrova, p. 37, pi. 18, fig. 10 [? = Hamulinites
parvulus (Uhlig)].

non 1967 Eoleptoceras (Tzankoviceras) assimilis (Uhlig); Nagy, p. 67, pi. 3, fig. 4 [? = Crioceratitinae juv.].

Holotype. BSM AS III 177; Upper(?) Barremian; Mistrovice, Silesian Unit, Czech Republic; figured by Uhlig
(1883’ pi. 32, fig. 9).

Revised diagnosis. Middle-sized hamulinicone, with simple sharp ribs crossing venter uninterrupted.
Interspaces twice as wide as ribs. On the phragmocone ribs project; on the possibly short living

VASICEK AND WIEDMANN: BARREM1AN AMMONITES

225

chamber ( = retroversum) ribs are radiate. Suture-line relatively complicated; lobes and saddles
with fine incisions.

Measurements. The fragmentary proversum measures 24 mm, Hmax = 6 mm. Length of retroversum is 19 mm,
with Hmax = 8 mm.

Remarks. A relatively large-sized leptoceratoid. Size and more complicated suture-line facilitate
distinction from other Hamulinites species. The species is similar to Anahamulina distans Vasicek
and 1 Hamulina ’ varusensis d’Orbigny, both of which have much more complicated suture-lines.

Specimens described from Bulgaria differ in ribbing, and approach H. parvulus (Uhlig). Revision
of the Hungarian material (Nagy 1967) leads to the conclusion that these forms are fine-ribbed
juveniles of ?Crioceratitinae; traces of suture-lines show strong denticulation.

Occurrence. Upper(?) Barremian, Mistrovice, Outer Western Carpathians, Czech Republic.

Genus veleziceras Wright, 1957

[= Orbignyceras Royo y Gomez, 1945 p. 462 (non Gerard and Contaut, 1936)].

Type species. Orbignyceras veleziensis Royo y Gomez, 1945. Barremian, Velez-Chipata, Colombia.

Revised diagnosis. Straight to gently curved shells with elliptical whorl section. Simple and projected
sharp ribs without any tubercle; eventually with faint periodic constrictions. Suture-line simple.

Remarks. Despite the fact that neither the early nor the late growth stages are known, a series of
forms with bochianitid coiling are included in Leptoceratoidinae. Inclusion is based on the
simplified suture-line: E is divided by a small median saddle, lobes are undivided, saddles are bifid.
This represents the most primitive (reduced?) leptoceratoid suture-line.

Veleziceras was previously (Wiedmann 1973) included in Karsteniceras , but may better be
considered distinct. It approaches in coiling Hamulinites , but has a much more complicated suture-
line. Besides the type species, V. hennigi (Stahlecker) is also included in Veleziceras.

Distribution. Barremian of Colombia and Maio, Cape Verde Islands.

Veleziceras veleziense (Royo y Gomez, 1945)

1945 Orbignyceras veleziensis Royo y Gomez, p. 462, pi. 71, fig. 1 d-e\ text-fig. 2.

1957 Veleziceras veleziense (Royo y Gomez); Wright, p. L210.

Lectotype (designated herein). SGNB specimen figured as Orbignyceras veleziense by Royo y Gomez (1945,
pi. 71, fig. \e)\ Barremian, Velez-Chipata, Colombia.

Diagnosis. As for the genus.

Remarks. Of the two specimens Royo y Gomez (1945) described as the 'holotype', the larger one is
here chosen as the lectotype.

Occurrence. Barremian of Velez-Chipata, Colombia.

Veleziceras hennigi (Stahlecker, 1935)
1935 Bochianites hennigi Stahlecker, p. 287, pi. 14, figs 18-20.

226

PALAEONTOLOGY, VOLUME 37

Lectotype (designated herein). GPIT 493/14/18; Barremian; Maio, Cape Verde Islands; figured by Stahlecker
(1935, pi. 14, fig. 18).

Revised diagnosis. Middle-sized, nearly straight shells with simple, slightly prorsiconcave ribs.
Width of ribs equal to interspaces. Suture-line unknown.

Remarks. This poorly known species is included in Veleziceras with some doubt. V. hennigi is similar
to the type-species, but can be separated by its more concave ribbing and the absence of
constrictions.

Occurrence. Barremian (presumably Lower) of Maio, Cape Verde Islands.

Genus eoheteroceras gen. nov.
Type species. Eoheteroceras silesiacum sp. nov.

Diagnosis. Small-sized ancylocones with strong simple ribs. Juvenile coil criocone-trochospiral.
Retroversum overlapping juvenile coil.

Description. First whorl unknown, juvenile whorl forming an indistinct trochospiral coil which can be covered
by the final retroversum. This makes clear differentiation and interpretation of the juvenile coil difficult. Adult
age ancylocone, with usually long retroversum. It can be assumed that the ancylocone stage underwent torsion
to avoid collision of the mouth border with the juvenile part. Due to sedimentary compaction in the complete
specimens, it is hard to estimate the distance between both parts. The shell is covered by simple, strong and
prorsiradiate ribs. Suture-line unknown.

Remarks. The juvenile stage separates Eoheteroceras gen. nov. from Hamulinites. Heteroceras
d'Orbigny can easily be separated by the distinctly helicoidal initial part and, in most cases, by
its larger size. Besides the type species, Eoheteroceras includes E. norteyi (Myczyriski and Trifif),
E. uhligi (Vasicek), and E. ? multicostatum (Stahlecker).

Distribution. Barremian of the Outer Carpathians of Poland and the Czech Republic; Southern Alps of
Switzerland; Prebetic of Spain; Cuba; and probably Maio, Cape Verde Islands.

Eoheteroceras silesiacum sp. nov.

Plate 4, figures 4-5

1883 Leptoceras n. f. ind.; Uhlig, p. 274, pi. 29, fig. 2.

Holotype. GBAW 3938; Lower(?) Barremian; Gorki Wielkie (Gurek), Poland (PI. 4, fig. 4); figured by Uhlig
( 1883, pi. 29, fig. 2).

Other material. GPIT 1719/9; Lower Barremian; La Querola, Spain.

Diagnosis. Small-sized elliptical shell with open criocone-trochospiral juvenile coil and ancylocone
adult hook. Retroversum long. Ribbing sharp and simple.

Description. Elliptically coiled small form with long retroversum and ribs as wide as interspaces.
Measurements. Dmax of the holotype = 23 5 mm, the deformed Hmax 10 5 mm.

VASICEK AND WIEDMANN: BARREMIAN AMMONITES 227

Remarks. Eoheteroceras norteyi is very similar to the present species, but its coiling is less elliptical
and the ribbing is finer. E. uhligi can be distinguished by its more involute juvenile coil and a much
shorter retroversum. E.l multicostatum is larger than the other species but its inclusion in the
present genus remains doubtful due to its very poor preservation.

Occurrence. Presumed Lower Barremian from Gorki Wielkie, Polish Outer Western Carpathians, and Querola,
Prebetic, Spain.

Eoheteroceras norteyi (Myczyriski and Trifif, 1986)

1986 Hamulinites norteyi Myczyriski and Triff, 1986, p. 127, pi. 1, fig. 8; pi. 3, fig. 6.

Holotype. MPMS-R AM-88/2; Barremian; Nortey, Pinar del Rio Province, Cuba; figured by Myczyriski and
Triff (1986, pi. 1, fig. 8).

Revised diagnosis. Small-sized ancylocone with criocone juvenile coil and long retroversum.
Interspaces much wider than sharp simple ribs.

Remarks. Despite the unknown suture-line, E. norteyi shows all of the characteristics of
Eoheteroceras. The mode of coiling contrasts with that of Hamulinites insofar as the retroversum
overlaps the criocone juvenile coil.

Occurrence. Barremian of Cuba.

Eoheteroceras uhligi (Vasicek, 1981)

Plate 4, figures 6-7

1883 Heteroceras n. f. ind.; Uhlig, p. 274, pi. 32, fig. 10.

1981 Heteroceras uhligi Vasicek, p. 123, pi. 1, fig. 2; pi. 2, fig. 2.

1990 Eoheteroceras uhligi (Vasicek); Vasicek, pi. 1, fig. 7.

Holotype. GBAW 3932, Lower(?) Barremian, Gorki Wielkie (not Jaworze, cf. Vasicek 1981, p. 123), Poland;
figured by Uhlig (1883, pi. 32, fig. 10).

Other material. VSB OT 5/30, Pi 1/6, OSM B 13 036; all crushed specimens; Lower Barremian; Outer
Carpathians of the Czech Republic; GPIT 1719/10, Upper Barremian of Breggia, Southern Alps.

Diagnosis. Small ancylocones with involute juvenile trochospiral stage, short proversum and final
hook. Strong and simple ribs are slightly prorsiradiate; interspaces become wider on retroversum.

Remarks. This species shows the typical shell morphology of Eoheteroceras. The initial coil is,
however, more involute than in the species described above, resembling therefore the younger
Heteroceras.

Occurrence. Lower Barremian of Ostravice, Trojanovice and Gorki Wielkie, Outer Western Carpathians of the
Czech Republic and Poland; Upper Barremian of Breggia Gorge, Ticino, Switzerland.

Eoheteroceras ? multicostatum (Stahlecker, 1935)
1935 Heteroceras multicostatum Stahlecker, p. 282, pi 13, figs 10-1 1.

228

PALAEONTOLOGY, VOLUME 37

U L E

text-fig. 6. Fragmentary suture-line of Mano/o-
viceras saharievae (Manolov) at H = 4-5 mm (after
Vasfcek 1972, fig. 19).

Lectotype (designated herein). GPIT 493/13/10; Lower(?) Barremian; Maio, Cape Verde Islands; figured by
Stahlecker (1935, pi. 13, fig. 10).

Revised diagnosis. Middle-sized species with elliptically criocone juvenile coil becoming arch-like
with age, and then overlapping the juvenile spiral portion. Final hook unknown. Ribbing strong
and simple.

Remarks. This species exhibits a certain degree of variation as shown by Stahlecker (1935). The
open juvenile trochospiral stage shows it to be more similar to Eoheteroceras gen. nov. than to
Heteroceras.

Occurrence. Lower(?) Barremian of Maio, Cape Verde Islands.

Genus manoloviceras gen. nov.

Type species. Hemibaculites saharievae Manolov, 1962; Lower Barremian; Kraptshene, Bulgaria.

Diagnosis. Small forms with low and open-whorled trochospiral initial coil becoming arch-like in
the adult. Ribbing fine at first, later on more accentuated. Suture-line simple.

Description. The initial coil of Manoloviceras gen. nov. forms an open low-angle trochospiral stage. Later on,
the shell becomes arch-like without forming a final hook. The juvenile shell is covered with fine and dense ribs
which become stronger with age. Suture-line is of leptoceratoid type with simple outlines.

Remarks. By comparison with this monospecific genus, the suture-line in Veleziceras is even more
simplified. Hemibaculites Hyatt is large, has a more complicated suture-line, and the shell closes up
with a final hook, which is also the case in Eoheteroceras gen. nov.

Distribution. Lower Barremian of Bulgaria, Hungary and the Czech Republic.

Manoloviceras saharievae (Manolov, 1962)

Plate 4, figure 8; Text-figure 6

1883 Hamites ( Anisoceras ) aff. obliquatum d'Orbigny ; Uhlig, p. 220.

1962 Hemibaculites saharievae Manolov, p. 536, pi. 73, figs 4—5, ?6.

1972 Veleziceras uhligi Vasfcek, p. 56, pi. 6, figs 1—2, text-figs 17-19.

1990 Eoheteroceras saharievae (Manolov); Vasfcek, pi. 1, fig. 5, pi. 2, figs 3, ?5.

229

VASICEK AND WIEDMANN: BARREMIAN AMMONITES

Holotype. SGM Su Crp 34; Lower Barremian; Kraptshene, Prebalkan, Bulgaria; figured by Manolov (1962,
pi. 73, figs 4-5).

Other material. OSM B 13 034, 13 039, 13 041, VSB OT 5/9, Pi 2/3, T 5/106, 267, 322; Outer Carpathians;
Czech Republic; one specimen, HNM Nagy collection, Gerecse Mts, Hungary.

Revised diagnosis. Initial coil indistinctly trochospiral, later on uncoiling in a simple curved arm
without final hook. Simple prorsiradiate ribs; ribs and interspaces covered with fine growth-lines.
Near the mouth, ribbing is totally replaced by these growth-lines. Suture-line (Text-fig. 6) simple,
with bipartite saddles, ceratite frilling and trifid U.

Remarks. Because of the uncommon uncoiling of M. saharievae , the species was previously included
in Hemibacidites , Veleziceras or even Eoheteroceras. Instead, a new genus of the Leptoceratoidinae
is here proposed for its reception. There is a certain similarity to the type species of Hemibacidites
Hyatt, H. obliquatus (d'Orbigny, 1842, pi. 120, figs 1^4) of which neither suture-line nor initial coil
are known. Yet, Hemibacidites is a much larger form, and the suture-line, at least of the North
American representatives such as H. mirabilis Anderson, is much more complicated.

Occurrence. Lower Barremian of Kraptshene, Prebalkan, Bulgaria ; Ticha and Shaft Frenstat 5, Pindula and
Ostravice, Silesian Unit, Outer Carpathians of the Czech Republic; Mt Berzsek, Gerecse Mts, Hungary.

SYSTEMATIC AND STR ATIGRAPHICAL CONCLUSIONS AND PHYLOGENY

It is obvious from the systematic descriptions that the recognition of members of the
Leptoceratoidinae is a difficult task. This is not only because of their small size but, above all, the
reduced outline of their suture-lines. Due to the predominant occurrence of leptoceratoid species in
clayey-shaly facies, preservation is, in many cases, rather poor and does not permit recognition of
the suture-line. Unfortunately, this is the case with many of the holotypes of proposed species;
juvenile, small-sized or fragmentary crioceratids, ancyloceratids, heteroceratids or hamulinids are
in consequence not easy to separate. However, in all these groups the differentiation and
denticulation of lobes and saddles is much greater at a comparable size (Wiedmann 1963;
Text-fig. 9). Due to these sutural differences, some of the previously attributed ‘leptoceratids’ have
to be excluded from this group, e.g. ‘ Leptoceras' varusense (d'Orbigny), ' L.' cirtae (Coquand) and
‘ L. ' puzosiamtm (d’Orbigny).

The combination of two prime characteristics (suture simplification and small size) suggests that
criocones, ancylocones, hamulinicones as well as hamiticones and even heteroceraticones should be
grouped in the same subfamily, Leptoceratoidinae, within the family Ancyloceratidae.

It can be seen from Text-figure 7 that, for stratigraphical reasons, Veveysiceras gen. nov. may
represent the source of this subfamily spreading in three divergent directions. Veveysiceras is known
from the External Prealps of Switzerland (Ooster 1860), from the Vocontian Trough and the Krizna
Nappe of the Western Carpathians of Slovakia. Here Veveysiceras is found immediately above the
beds with Pseudothurmannia which implies an early Lower Barremian age. A first occurrence within
the Hauterivian together with Pseudothurmannia is assumed.

In the Krizna Nappe of the Western Carpathians the first members of Hamulinites co-occur with
Veveysiceras. Hamulinites , representing the first evolutionary line in the Leptoceratoidinae (Text-
fig. 7), becomes abundant and widespread in the middle and upper part of the Lower Barremian.
While the small-sized H. Jragilis is rare and geographically restricted (External Prealpes, Western
Carpathians), H. parvulus is abundant and cosmopolitan during the Lower Barremian. H. assimilis
is considered to be the last member of this genus, known only from the lower Upper Barremian of
the Outer Carpathians.

Veleziceras Royo y Gomez is one of the rather poorly known representatives of this subfamily.

230

PALAEONTOLOGY, VOLUME 37

a

ed

© v=\

6

<D

J-i

5-h

ed

PQ

<L>

a

Oi

p

a

ed

S

5-1

5-1

ed

PQ

5-i

<D

£

O

P

Heteroceras

astieri

Manoloviceras

saharievae

Eoheteroceras

silesiacum

Karsteniceras

subtile

fragilis

text-fig. 7. Evolution of Barremian Leptoceratoidinae.

VASICEK AND WIEDMANN: BARREMIAN AMMONITES

231

The type species, V. orbignyi, is known from Colombia, while a second species, V. hennigi
Stahlecker, was described from Maio, Cape Verde Islands. Both species are considered to be Lower
Barremian. The origin of Veleziceras remains uncertain.

Karsteniceras constitutes the central stock of the subfamily. It comprises the most diverse and the
longest ranging group of forms and is defined by its weakly trochospiral and criocone uncoiling.
Karsteniceras appears with K. subtile and K. pumilum in the early Lower Barremian of Europe.
Endemic regional offshoots are widespread during the Lower Barremian, i.e. the Japanese
K. asiaticum and K. obatai , the Caribbean K. trinidadense and K. polieri , and the Cape Verdean
K.l filicostatum. In the upper Lower and lower Upper Barremian a number of species develop
marginal tubercles (K. beyrichi , K. beyrichioide , K. ibericum , K. hoheneggeri ) or a different type of
ribbing on the living chamber. These species are more widespread than the above-mentioned species
and may define a number of subprovinces.

Another evolutionary line evolves from Veveysiceras towards Eoheteroceras and Manoloviceras.
Eoheteroceras represents the ‘type’ of this evolutionary line and to some extent foreshadows the
later genus Heteroceras. It is long-ranging, starting in the early Lower Barremian with the
Carpathian species E. silesiacum ; including in the Lower Barremian the endemic E. norteyi (from
Cuba) and E. ? multicostatum (from the Cape Verde Islands); and ending up with E. uhligi , which
is known from the late Lower Barremian of the Outer Carpathians and early Upper Barremian of
Ticino, Southern Alps. Increasing size seems to be a general trend in each of the three separate
evolutionary lines. Manoloviceras , however, is considered to be a lateral offshoot loosing the final
hook. It is regionally rather restricted and of Lower Barremian age.

Text-figure 8 gives an idea of suture-line constancy and variation in leptoceratids. At first glance,
suture-lines of the main representatives of this group look very similar with regard to their
simplification in the adult stage. However, the three evolutionary lines defined above (Text-fig. 7)
can be recognized by specific lobe or saddle differentiation. While in Hamulinites saddles tend to be
more frilled, in Manoloviceras the frilling of lobes increases. The cryptogenous Veleziceras exhibits
the most simplified adult suture-line (Text-fig. 9).

While we can assume that the evolutionary line Veveysiceras-Eoheteroceras leads to the large-
sized Heteroceras d'Orbigny near the Lower/Upper Barremian boundary, we have no idea yet
about the ancestor of either Veveysiceras or the Leptoceratoidinae. These forms are certainly
unrelated to the homoeomorphic true leptoceratids of Berriasian to Valanginian age.

PALAEOGEOGRAPHY

Text-figures 10 and 1 1 give the regional distribution of leptoceratoids. The Leptoceratoidinae have
their main distribution on the southern shelf margin of the European Plate, from the Caucasian
Basin in the east, to the Prebetic and Subbetic Basins in southern Spain in the west.

Akopyan ( 1962) described Karsteniceras cf. pumilum from the Zangezur Mts, Armenian Caucasus
(Kafan area of Kotetishvili 1989). Favoured areas for leptoceratoids were the Bulgarian Prebalkan
and the southern and western margins of the Moesian Platform. From Bulgaria, Lower Barremian
Hamulinites, Karsteniceras and Manoloviceras were recorded by Manolov (1962), Breskovski
(1966), Dimitrova (1967) and Nikolov (1987), from a well oxidized marly facies. Avram (1988)
summarized the Romanian occurrences. Here, all the Bulgarian genera were also recorded from
the western Moesian Platform margin, e.g. mainly from pelagic facies at Svinffa, Marginal Dacides.
From flysch deposits of the Outer Dacidic Nappes, Hamulinites and Karsteniceras were reported,
while from the pelagic facies in the Getic Nappe (Median Dacides) only the genus Karsteniceras is
known.

Another evolutionary centre for leptoceratoids during Lower Barremian (and lower Upper
Barremian) time was the Silesian Nappe in the Outer Western Carpathians of the Czech Republic
and Poland (Vasicek 1972, 1981, 1990). In dark-grey flyschoid clays, ten species can be detected
which belong to Karsteniceras , Hamulinites, Eoheteroceras and Manoloviceras in descending order.

232

PALAEONTOLOGY, VOLUME 37

E L U

Veveysiceras

text-fig. 8. Evolutionary trends in adult suture-lines of the Leptoceratoidinae.

The next occurrences further west are the Austrian Helvetic Drusberg Marls with Karsteniceras
(Wiedmann 1977), and the dark-grey marls of the External Prealps (Vaud, Switzerland). In
Switzerland, nearly the same diversity as in the Carpathians was detected by Ooster (1860) and
Sarasin and Schondelmeyer (1902). The Swiss leptoceratoids are here referred to the genera
Karsteniceras , Hamulinites and Veveysiceras.

From the southern margin of the Vocontian Trough, southeastern France, Veveysiceras and
Hamulinites are newly described from light-grey Lower Barremian marls. The westernmost
occurrences in the European Tethyan Realm are reported in light-grey Lower Barremian marls from
the Prebetic Platform and the Subbetic Trough in southern Spain.

Leptoceratoidinae occur less abundantly on the northern margin of the African Plate.
Karsteniceras and Eoheteroceras are recognized (Rieber 1977; this paper) in Lower-Upper
Barremian dark intercalated shales in the upper Maiolica in Ticino, Southern Alps. Karsteniceras
and Hamulinites occur in a light-grey marly facies in the Lower Barremian Rossfeld Beds of the
Austroalpine Alps.

An easternmost occurrence, light-grey slope marls of the Krizna Nappe, Central Western
Carpathians of Slovakia, has yielded Veveysiceras, Karsteniceras and Hamulinites (Borza et al.
1984; this paper).

Considering the global distribution of the Leptoceratoidinae (Text-fig. 10), the supposed
occurrences in northern Africa cannot be confirmed, i.e. ‘ Leptoceras' cirtae Coquand, 1880 from
Algeria, and the ‘ Leptoceras ' puzosianum (d’Orbigny) of Burrolet et al. (1983) from Tunisia.

233

VASICEK AND WIEDMANN: BARREMI AN AMMONITES

text-fig. 9. Comparison of suture ontogenies in leptoceratoids. a, Hamulinites parvulus (Uhlig) [= H. munieri
(Nickles)], GPIT 1247/2; Lower Barremian; La Querola, Cocentaina, Spain; a, prosuture; b, primary suture
and septal surface; c, at WH 0-5 mm; d, intermediate stage; e, at WH 2-5 mm (from Wiedmann 1963).
b, Leptoceras studeri (Ooster), GPIT 1372/2; Berriasian; Rufisgraben-Jusital, Switzerland; a, primary suture;
b, tenth suture and septal surface; c, 13th suture; d, 20th suture at WH L5mm; e, intermediate stages;
h, adult suture at WH 3-5 mm (from Wiedmann 1969).

The equatorial-subequatorial Atlantic was, however, another area favoured by leptoceratoids. In
the SE North Atlantic, Stahlecker (1935) discovered representatives of the genera Karsteniceras ,
Veleziceras and Eoheteroceras in platy basinal Lower Barremian limestones from Maio, Cape Verde
Islands. Another centre of occurrence is the Caribbean. Here, we know Karsteniceras from Trinidad
(Imlay 1954) and Hamulinites and Eoheteroceras in micritic limestones of the Polier Formation of
Cuba (Myczyriski 1977; Myczyriski and Triff 1986). Karsteniceras is described from the Lower
Barremian Santuario Formation in Queretaro, Mexico, (Gonzalez-Arreola and Carillo-Martinez
1986), while the black nodular Lower Barremian limestones of the Sabana de Bogota, Colombia,
contain Karsteniceras and Veleziceras (Karsten 1858; Royo y Gomez 1945; Etayo Serna 1968).
Recent investigations in the Japanese Lower Cretaceous by Matsukawa (1987, 1988) have added the
western Pacific to the distributional range. Karsteniceras was recorded from the clayey-sandy Ishido
Formation and concretions in the Kimigahama Formation, both in the Lower Barremian of
Hondo.

The spreading and evolutionary centre of the Leptoceratoidinae was thus the northern margin of
the western and central Tethys (Text-fig. 11). Especially on the southern shelf of the central
European Plate leptoceratoids prospered in basinal environments with a dominance of clayey-marly
facies. From this evolutionary centre, they presumably followed the North Equatorial Current in

234

PALAEONTOLOGY, VOLUME 37

text-fig. 10. Distribution of Barremian Leptoceratoidinae in Europe. Endemic species are not considered. Palaeogeography adopted from Dercourt et
a!. (1985, 1990). Abbreviations: B, Briangonnais; BS, Black Sea; Bii. Biikk, Cs, Caspian Sea; CSH, Cor-Sardinian High; Cz, Czorsztyn; Da. Danubian;
EP, Eastern Pontides; G, Getic; H, Helvetic; IBM, Iberian Meseta; MC, Central Massif; Me, Mecsek; Mo, Moesia; NC, Northern Calcareous Alps;
NT, Northern Transcaucasus; P, Piemontais; RBM, Rhenish-Bohemian Massif; Ro, Rossfeld; Si, Silesian; Sp, South Penninic; ST, Southern
Transcaucasus; Ta, Tatnc; US, Ukrainian Shield; Vi, Villany; WP, Western Pontides; Yb, Ibbsitz.

235

VASICEK AND WIEDMANN: BARREMIAN AMMONITES

▲ Veveysiceras • Eoheteroceras ■ Veleziceras

A Manoloviceras O Karsteniceras □ Hamulinites

text-fig. 11. Global distribution of Leptoceratoidinae; Barremian palaeogeography after Barron (1987).

a westward direction (see also Gordon 1973; Klinger et cil. 1984; Obata and Matsukawa 1988;
Klinger 1990) to colonize the Atlantic and Pacific oceans.

PALAEOECOLOGY

Leptoceratoid palaeoecology is highly speculative but can be inferred using the following data.

Leptoceratoids are to some extent global in distribution, but have an interesting E-W gradient
ranging from the evolutionary centre in central and southeastern Europe to Japan. In the western
Tethys they are practically restricted to the northern oceanic margin. The favoured sedimentary
facies are clayey and limey marls. These marls are either dark-grey to black with increasing contents
of organic matter and pyrite, or light-grey and well-oxygenated. Leptoceratoids also occur in flysch
and flysch-like deposits.

Leptoceratoidinae show variation in shell size, shell form, mode of uncoiling, strength and density
of ribbing, and in the otherwise very stable form of the external lobe E. This may be related to a
more or less distinct shell torsion in some of the genera.

The only discussion of life habit available is given by Rieber (1977), who described Karsteniceras
balernaense from Breggia. He concluded that this species led a nektonic existence in the oxygenated
Tethyan Ocean above anoxic bottom conditions. These conditions were inferred from the black
marly sediment enriched in organic matter and pyrite, and from the absence of definite benthos.

Westermann ( 1 990), in a stimulating paper on ammonite depth distributions, located the criocone
heteromorphs in shallow marine habitats of the inner shelf.

One of the most abundant occurrences of leptoceratoids is in the dark-grey marls of the Tesin-
Hradiste Formation (formerly Wernsdorfer Beds) of the Silesian Unit, Outer Carpathians (Uhlig
1883; Vasicek 1977). These beds (and most probably their organic content too) are, however, of
turbiditic origin. Leptoceratoids have not only their maximum diversity, but also their maximum
density within these beds, where they sometimes occur concentrated in 'nests’ (PI. 4, fig. 8). These
maxima are generally found in the autochthonous layers between two turbidites. Perfect
preservation, especially of the fragile initial coil and the co-occurrence of two-valved inoceramids

236

PALAEONTOLOGY, VOLUME 37

(PI. 2, fig. 6), supports the idea that shells were not transported but were deposited close to the
original biotope. It is true, however, that the total faunal ‘association’ is diverse and includes
phylloceratids and lytoceratids as well as heteroceratids and true Ammonitina. This association is
considered to be a representative cross-section through the whole water column.

These observations have much in common with those of Dietl (1973, 1978) who described Middle
Jurassic spiroceratids having their maximum distribution in the Suebian basinal facies with stagnant
conditions. Spiroceratid abundance decreases towards the more oxygenated oolitic shallow water
realm. Dietl (1978, p. 70) inferred a bottom-related, vagrant life habit for spiroceratids which
preferred to live in algal mats. The similar depositional environment and shell morphology (partial
torsion) of spiroceratids and leptoceratoids support the idea of a similar life habit (Text-fig. 12).

LOWER BARREMIAN

text-fig. 12. Model of leptoceratoid life habit.

From their geographical distribution an E W gradient can be observed, while the Tethyan Ocean
was an obvious barrier for N-S crossing. Transportation by the E-W running North Equatorial
Current is unlikely for benthic animals. As most cephalopods favour nearshore breeding places,
leptoceratoids may have been prone to drifting in surface currents during their early juvenile stages.
For these tiny and fragile forms, however, a planktonic mode of life cannot be totally refuted.

Acknowledgements. Both authors are much indebted to the Alexander von Humboldt Foundation for
facilitating this collaboration, to W. Wetzel, K. Mezihorakova, M. Grmelova and D. Korn tor the good
quality of the photographs, to Dr F. Etayo Serna (Bogota), Dr W. Kuhnt (Tubingen) and J. Manthey
(Tubingen) for contributing additional specimens, and to Dr D. Decrouez (Geneve), Dr H. Immel (Miinchen),
Dr I. Z. Nagy (Budapest), Dr G. Schairer (Miinchen), Dr H. A. Stalder (Bern) and Dr F. Stojaspal (Wien)

VASICEK AND WIEDMANN: B A RREMI AN AMMONITES

237

for making available specimens and casts from their collections. Some of the figures were kindly prepared by
H. Vollmer and R. Schoch (Tubingen). The authors appreciate the valuable comments of the reviewers.

REFERENCES

akopyan, v. T. 1962. Stratigrafiya yurskikh i myelovykh otlosheniy yugo-vostotshnovo Zangezura. Izdatyelstvo
Akademiya Nauk Armyanskoi SSR, Erevan, 265 pp. [In Russian].
astier, J. e. 1851. Catalogue des Ancyloceras appartenant a l'etage neocomien d’Escragnolles et des Basses-
Alpes. Annales des Sciences Physiques et Naturelles, cf Agriculture et de F Industrie, 2eme serie, 3, 435-456.
avram, E. 1976. La succession des depots Tithoniques superieurs et Cretaces inferieurs de la region de Svini|a
(Banat). Dari Searnci ale §edinjelor, 62. 53-71.

— 1988. The early Cretaceous (Berriasian-Barremian) ammonite assemblages in Roumania, 607-619. In

wiedmann, j. and kullmann, j. (eds). Cephalopods - present and past. Schweizerbart, Stuttgart, xiii + 765

pp.

— and kusko, m. 1984: Cephalopodes eocretaces de la partie centrale et meridionale des monts Baraolt
(Carpathes orientales). Dari Seamd ale § edin(elor , 79, 5-24.

barron, e. j. 1987. Global Cretaceous paleogeography - International Geologic Correlation Program Project
191. Palaeogeographv , Palaeoclimatology, Palaeoecology , 59, 207-214.
borza, K., MiCHALiK, J., gasparikova, v. and vasicek, z. 1984. The biostratigraphy of the Hauterivian/
Barremian boundary beds in the Krizna Nappe, Western Carpathian (Czechoslovakia). Cretaceous
Research , 5, 349-356.

breskovski, s. 1966. Biostratigraphie du Barremian au sud du village de Brestak, dans la region de Varna.
Travaux sur la Geologie de Bulgarie , 8, 71-1 12.

burollet, p. f., memmi, l. and m’rabet, a. 1983. Le Cretace inferieur de Tunisie. Aperqu stratigraphique et
sedimentologique. Zitteliana , 10, 255-264.

coquand, h. 1880. Etudes supplementaires sur la paleontologie algerienne, faisant suite a la description
geologique de la region sud de la province de Constantine. Bulletin de F Academic Hippone , 15, 1^149.
cottreau, J. 1937. Types du Prodrome de Paleontologie stratigraphique universelle de D'Orbigny. Tome III
(Neocomien). Part 3. Annales de Paleontologie , 26, 53-84.
darga, r. and weidich, k. F. 1986. Die Lackbach-Schichten, erne klatische Unterkreide-Serie in der Unkener
Mulde (Nordliche Kalkalpen, Tirolikum). Mitteilungen der Bayerischen Staatssammlung fur Paldontologie
und Historische Geologie , 26, 93-109.

dercourt, j., and 16 others. 1985. Presentation de 9 cartes paleogeographiques au 1 /20.000.000e s'etendanl
de l’Atlantique au Pamir pour la periode du Lias a l’Actuel. Bulletin de la Societe Geologique de France , 8eme
serie, 1, 637-652, 10 pis.

— RICOU, L. E., ADAMIA, S., CZASZAR, G., FUNK, H., LEFELD, J., RAKUS, M., SANDULESCU, M„ TOLLMANN, A. and
tchoumachenko, p 1990. Anisian to Oligocene paleogeography of the European margin of Tethys (Geneva
to Baku). 159-190, 1 1 maps. In rakus, m., dercourt, j. and nairn, a. e. m. (eds). Evolution of the northern
margin of Tethys : The results of IGCP Project 198. Volume 3. Memoires de la Societe Geologique de France ,
154 (3), 1-220, 11 maps.

dietl, G. 1973. Middle Jurassic (Dogger) heteromorph ammonites. 283-285. In hallam, a. (ed . ). Atlas of
palaeobiogeography. Elsevier, Amsterdam, xii + 531 pp.

— 1978. Die heteromorphen Ammoniten des Dogger (Stratigraphie, Taxonomie, Phylogenie, Okologie).
Stuttgarter Beitrage fur Naturkunde , 33, 1-76.

Dimitrova, N 1967. Les fossiles de Bulgarie. IV. Cretace inferieur. Cephalopoda (Nautiloidea et Ammonoidea).
Academie Bulgare des Sciences, Sofia, 236 pp. [In Bulgarian with French summary].

1970. Phylogenese des Ammonites heteromorphes du Cretace inferieur. Izvestiya na Geologitsheskaya
Institute Seriya palaeontologiya , 19, 71-1 10.

drushtchic, v. v. 1960. [Ammonites Part 1]. 249-308. In drushtchic, v. v. and Kudryavtsev, m. p. (eds).

Atlas nishnemyelovoy fauny severnovo Kavkaza i Kryma. Gostoptekhyzdat, Moscow, 396 pp. [In Russian].
etayo serna, F. 1968. Apuntaciones acerca de algunas amonitas interesantes del Hauteriviano y del
Barremiano de la region de Villa de Leiva (Boyaca, Colombia, S.A.). Boletin Geologico de la Universidad de
Santander, 24, 51-70.

gerard, c. and contaut, h. 1936. Les ammonites de la zone a Peltoceras athleta du Centre-Ouest de la France.

Memoires de la Societe geologique de France, Nouvelle Serie, 13, memoire 29, 79 pp.
gill, t. 1871. Arrangement of the families of mollusks. Smithsonian Miscellaneous Collections, 227,
i-xvi + 1-49.

238

PALAEONTOLOGY, VOLUME 37

gonzalez-arreola, c. and CARiLLO-MARTiNEZ, m. 1987. Amonitas del Jurasico superior (Titoniano superior) y
del Cretacico inferior (Hauteriviano-Barremiano) del area de San Joaquin-Vizarron, Estado de Queretaro.
Universidad National Autonoma de Mexico, Instituto de Geologia, Revista, 6 (2), 171-177.

Gordon, w. a. 1973. Marine life and ocean surface currents in the Cretaceous. Journal of Geology, 81, 269-284.
imlay, R. w. 1954. Barremian ammonites from Trinidad, B.W.I. Journal of Paleontology, 28, 662-667.
immel, h. 1987. Die Kreideammoniten der nordlichen Kalkalpen. Zitteliana, 15, 3-163.
karsten, h. 1858. Uber die geognostischen Verhaltnisse des westlichen Columbien, der heutigen Republiken
Neu-Granada und Equator. Amtliche Bericlite, 32, Versammlung Deutscher Naturforscher und Ante, 80-1 17.
kilian, w. 1907-1913. Unterkreide (Palaeocretacium). 398 pp. In frech, f. (ed.) Lethaea geognostica. II.

Mesozoikum, Volume 3. Kreide. Schweizerbart, Stuttgart.
klinger, h. c. 1990. Upper Barremian Heteroceratinae (Ammonoidea) from the Caucasus, Zululand and
Patagonia: palaeobiogeographic significance. Cretaceous Research 11, 321-329.

kakabadze, m. v. and Kennedy, w. j. 1984. Upper Barremian (Cretaceous) heteroceratid ammonites
from South Africa and the Caucasus and their paleobiogeographic significance. Journal of Molluscan
Studies , 50, 43-60.

kotetishvili, e. v. 1989. Biofacies characteristics of Lower Cretaceous deposits of Georgia. 543-550. In
wiedmann, J. (ed.). Cretaceous of the Western Tetliys. Schweizerbart, Stuttgart, xiv+1005 pp.
kullmann, j. and wiedmann, j. 1970. Significance of sutures in phylogeny of Ammonoidea. University of
Kansas Palaeontological Contribution, 47, 1-32.

luppov, N. p. and drushtchic, v. v. 1958. Mollyuski - Golovonogiye. IE Ammonoidei (Ceratity i Ammonity).
Vnutrennerakovinnye. Prilozheniye : Konikonkhii. Osnovy Paleontologii. Gosudarstvennoye N.T. Izdatyel-
stvo, Moscow, 190 pp. [In Russian],

manolov, j. r. 1962. New ammonites from the Barremian of North Bulgaria. Palaeontology, 5, 527-539.
manthey, j. 1991 Stratigraphische und fazielle Untersuchungen im westlichen Subbetikum (Sierra de las
Cabras, Sierra de la Sal und Sierra del Valle, Provincia de Cadiz, Siid-Spanien). Unpublished Diplome
Thesis, Universitat Tubingen.

matsukawa, m. 1987. Early shell morphology of Karsteniceras (ancyloceratid) from the Lower Cretaceous
Choshi Group, Japan and its significance to the phylogeny of Cretaceous heteromorph ammonites.
Transactions and Proceedings of the Palaeontological Society of Japan, New Series , 148, 346-359.

1988. Barremian ammonites from the Ishido Formation, Japan - supplements and faunal analysis.
Transactions and Proceedings of the Palaeontological Society of Japan , New Series , 149, 396-416.
myczynski, r. 1977. Lower Cretaceous ammonites from Sierra del Rosario (Western Cuba). Acta
Palaeontologica Polonica, 22, 139-173.

— and triff, j. 1986. Los ammonites del Cretacico Inferior de las provincias de Pinar del Rio y Matanzas.
Bulletin of the Polish Academy of Sciences , Earth Sciences, 34, 113-137.

nagy, i. z. 1967. Unterkretazische Cephalopoden aus dem Gerecse-Gebirge. Annales Historico-Naturales
Musei Nationalis Hungarici, pars Miner alogica et Geologica, 59, 53-73.
nickles, r. 1894. Contributions a la paleontologie du Sud-Est de l’Espagne. Terrain Cretace. Neocomien
(suite). Memoires de la Societe Geologique de France, 4 (Memoire 4), 31-59.
nikolov, t. 1967. Protoleptoceras gen. n. - a new genus of Berriassian ammonites. Comptes Rendus de
I" Academie Bulgare des Sciences, 19 (9), 839-841.

obata, i. and matsukawa, m. 1988. Some boreal or subboreal ammonites in the Japanese Barremian. 469-476.
In wiedmann, j. and kullmann, j. (eds). Cephalopods - present and past. Schweizerbart, Stuttgart, xiii + 765
PP-

ooster, w. a. 1860. Catalogue des Cephalopodes fossiles des Alpes suisses. IV. Partie: Cephalopodes
tentaculiferes, Ammomtides, G. Ammonites. Neue Denkschriften der Allgemeinen Schweizerischen Gesell-
schaft der Gesamten Naturwissenschaften, 18. 160 pp.

orbigny, a. d\ 1840-1842. Terrains cretaces. I. Cephalopodes. Paleontologie frangaise. Masson, Paris, 662 pp.

— 1 850. Prodrome de Paleontologie stratigraphique universelle des animaux mollusques et rayonnes. Corallien-
Cretace-Parisien, 2. Masson, Paris, 427 pp.

— 1852. Notice sur le genre Hamulina. Journal de Conchy liologie, 3, 207-228.

paquier, v. 1900-1901 . Recherches geologiques dans le Diois et les Baronnies orientales. Annales de f Universite
de Grenoble, 12, 373-516 + appendix.

pictet, f.-j. and loriol, p. de 1858. Description des fossiles contenu dans le terrain Neocomien de Voirons,
Description des animaux invertebres. Materiaux de la Paleontologie Suisse, 2(1), 1-64.
rieber, h. 1977. Eine Ammonitenfauna aus der oberen Maiolica der Breggia-Schlucht (Tessin, Schweiz).
Eclogae geologiae Helvetiae, 70, 777-787.

239

VASICEK. AND WIEDMANN: BARREMI AN AMMONITES

roman, F. 1938. Les ammonites Jurassiques et Cretacees. Essai de Genera. Masson, Paris, 554 pp.
royo y gomez, j. 1945. Fosiles del Barremiense colombiano. Compilaciones de estudios geolbgicos oficiales de
Colombia , 6, 455^195.

sarasin, ch. and schondelmayer, c. 1902. Etude monographique des ammonites du Cretacique inferieur de
Chatel-Saint-Denis. Memoires de !a Societe Paleontologique Suisse , 28/29, 195 pp.
sarkar, s. 1954. Some new genera of uncoiled ammonites from Lower Cretaceous. Science and Culture , 19 (12),
618-620.

1955. Revision des ammonites deroulees du Cretace inferieur du Sud-Est de la France. Memoires de la
Societe Geologique de France , Nouvelle Serie , 72, 176 pp.
sayn, G. 1891. Description des ammonites du Barremian du Djebel-Ouach (pres Constantine). Annales de la
Societe if Agriculture , 6eme serie, 3, 135-208.

stahlecker, r. 1935. Neocom auf der Kapverden-Insel Maio. Neues Jahrbuch fur Mineralogie , Geologie und
Palaontologie , 1934, 265-301.

thieuloy, j.-p. 1966. Leptoceres berriasiens du massif de la Grande-Chartreuse. Travaux du Laboratoire de
Geologie , Grenoble , 42, 281 -295.

uhlig, v. 1883. Die Cephalopoden der Wernsdorfer Schichten. Denkschriften der k. k. Akademie der
Wissenschaften Wien , Mathematisch-Naturwissenschaftliche Klasse , 46, 127-290.
vasicek, z. 1972. Ammonoidea of the Tesin-Hradiste Formation (Lower Cretaceous) in the Moravskoslezke
Beskydy Mts. Rozpravy Ustredmho Ustavu Geologickeho, 38, 103 pp.

1981. Zwei neue faunistische Fundorte der Unterkreide in der Silesischen Einheit und Zusammenfassung
der Revision der von V. Uhlig (1883) beschriebenen Ammoniten. Sbornik Vedeckych Praci Vysoke Skoly
Bdnsske v Ostravse, 25 (2, for 1979), 119-133.

1990. Unterkreide-Ammomten aus neu abgeteuften Schacheten im Gebiet von Frenstat-Trojanovice
(Aussere Karpaten, Silesische Einheit, CSFR). Acta Musei Moraviae , Scientiae Naturales, 75, 95-106.
Wedekind, r. 1916. Zur Systematik der Ammonoidea. Centralblatt fur Mineralogie , 1916, 529-538.
westermann, G. e. G. 1990. New developments in ecology of Jurassic-Cretaceous ammonoids. 459-478. In
pallini, g., cecca, f., cresta, s. and sant antonio, m. (eds). Fossili, Evoluzione , Ambiente , Atti del 2°
Convegno internazionale F.E.A. , Pergola 1987. Comitato Centenario Raffaelo Piccinini, Pergola, 516 pp.
wiedmann, j. 1963. Entwicklungsprinzipien der Kreideammoniten. Palaeontologische Zeitschrift , 37, 103-121.
1969. The heteromorphs and ammonoid extinction. Biological Reviews , 44, 563-602.

— 1972. Neue Vorstellungen liber Stammesgeschichte und System der Kreideammoniten. Proceedings of the
23rd International Geological Congress. Prague 1968 , International Palaeontological Union , 93-118.

1973. Ancyloceratina (Ammonoidea) at the Jurassic/Cretaceous boundary. 309-316. In hallam, a. (ed.).
Atlas of palaeobiogeography. Elsevier, Amsterdam, xii + 531 pp.

— 1977. On the significance of ammonite nuclei from sieve residues. Annales des Mines et de la Geologie ,
28, 135-161.

wright, c. w. 1957. Cretaceous ammonites. In arkell, w. j., kummel, b. and wright, c. w. Mesozoic
Ammonoidea. L80-L465. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Part L. Mollusca 4.
Cephalopoda-Ammonoidea. Geological Society of America and University of Kansas Press, New York and
Lawrence, Kansas, 490 pp.

1981. Cretaceous Ammonoidea. 157-174. In house, m. r. and senior, j. r. (eds). The Ammonoidea.
Systematics Association Special Volume, 18. Academic Press, London, xiv + 593 pp.
yabe, h., nagao, t. and shimizu, s. 1926. Cretaceous Mollusca from the Sanchu Graben in the Kwanto
Mountainland, Japan. Science Report of the Tohoku Imperial University , Series 2 , Geology , 9, 33-76.

ZDENEK VASICEK

Department of Geology and Mineralogy
Mining University
Tr. 1 7. listopadu

CR-70833 Ostrava-Poruba, Czech Republic

JOST WIEDMANN

Geologisches und Palaontologisches Institut
Universitat Tubingen

Typescript received 1 September 1992 Sigwartstrasse 10

Revised typescript received 6 July 1993 D-7400 Tubingen, Germany

NOTES FOR AUTHORS

The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts should conform
in style to those already published in this journal, and should be sent to Dr R. M. Owens, Department of Geology, National
Museum of Wales, Cardiff CF1 3NP, UK, who will supply detailed instructions for authors on request (these are published
in Palaeontology 1990, 33, pp. 993-1000).

Special Papers in Palaeontology is a series of substantial separate works conforming to the style of Palaeontology.

SPECIAL PAPERS IN PALAEONTOLOGY

In addition to publishing Palaeontology the Association also publishes Special Papers in Palaeontology. Members may
subscribe to this by writing to the Membership Treasurer: the subscription rate for 1994 is £55 00 (U.S. $120) for Institutional
Members, and £20 00 (U.S. $36) for Ordinary and Student Members. A single copy of each Special Paper is available on a
non-subscription basis to Ordinary and Student Members only , for their personal use, at a discount of 25 per cent below the
listed prices: contact the Marketing Manager. Non-members may obtain Nos 35^16 (at cover price) from Basil Blackwell Ltd.,
Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, UK, and older issues from the
Marketing Manager. For all orders of Special Papers through the Marketing. Manager, please add £ 1 50 (U.S. $3) per item
for postage and packing.

PALAEONTOLOGICAL ASSOCIATION PUBLICATIONS

Special Papers in Palaeontology

For full catalogue and price list, send a self-addressed, stamped A4 envelope to the Marketing Manager. Numbers 2-39 are
still in print and are available together with those listed below :

39. (for 1988): Late Cenomanian and Turonian ammonite faunas from north-east and central Texas, by w. j. Kennedy,
131 pp.. 39 text-figs , 24 plates. Price £30 (U.S. $60).

40. (for 1988): The use and conservation of palaeontological sites. Edited by p. r. crowther and w. a. Wimbledon. 200 pp..
31 text-figs. Price £30 (U.S. $60).

41. (for 1989): Late Jurassic-early Cretaceous cephalopods of eastern Alexander Island, Antarctica, by p. j. howlett.
72 pp.. 9 text-figs, 10 plates. Price £20 (U.S. $40).

42. (for 1989): The Palaeocene flora of the Isle of Mull, by m. c. boulter and z. kvacek. 149 pp., 23 text-figs, 23 plates.
Price £40 (U.S. $80).

43. (for 1990) : Benthic palaeoecology of the Late Jurassic Kimmeridge Clay of England, by p. r. wignall, 74 pp., 50 text-figs.
Price £30 (U.S. $60).

44. (for 1990): The Murchison Symposium. Edited by M. G. bassett, p. d. lane and d. Edwards. 397 pp-., 137 text-figs,
3 plates. Price £60 (U.S. $120)

45. (for 1991): Coniacian ammonite faunas from the United States Western Interior, by w. J. Kennedy and w. A. Cobban.
96 pp., 31 text-figs, 17 plates. Price £35 (U.S. $70).

46. (for 1991): Arundian (Lower Carboniferous) conodonts from South Wales, by J. I. stone. 63 pp., 14 text-figs, 5 plates.
Price £30 (U.S. $60)

47. (for 1992): The skeletal morphology, systematics and evolutionary history of holothurians, by p. m. Gilliland. 147 pp.,
19 text-figs, 11 plates.

48. (for 1992): Contributions to acritarch and chitinozoan research. Edited by s. G. molyneux and k. j. dorning. 139 pp.,
28 text-figs, 22 plates.

49. (for 1993): Studies in palaeobotany and palynology in honour of Professor W. G. Chaloner, F.R.S. Edited by
m. e. collinson and a. c. scott. 187 pp., 38 text-figs, 27 plates.

Field Guides to Fossils and Other Publications

These are available only from the Marketing Manager. Please add £100 (U.S. $2) per book for postage and packing plus
£1 50 (U.S. $3) for airmail. Payments should be in Sterling or in U.S. dollars, with all exchange charges prepaid. Cheques
should be made payable to the Palaeontological Association.

1. (1983): Fossil Plants of the London Clay, by m. e. collinson. 121 pp., 242 text-figs. Price £7-95 (U.S. $16) (Members £6

or U.S. $12).

2. (1987): Fossils of the Chalk, compiled by e. owen; edited by a. b. smith. 306 pp., 59 plates. Price £11 50 (U.S. $23)

(Members £9-90 or U.S. $20).

3. (1988): Zechstein Reef fossils and their palaeoecology, by n. hollingworth and t. Pettigrew, iv + 75 pp. Price £4-95

(U.S. $10) (Members £3/75 or U.S. $7-50).

4. (1991): Fossils of the Oxford Clay, edited by d. m. martill and j. d. Hudson. 286 pp., 44 plates. Price £15 (U.S. $30)

(Members £12 or U.S. $24).

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $40).

1985. Atlas of Invertebrate Macrofossils. Edited by j. w. Murray. Published by Longman in collaboration with the
Palaeontological Association, xiii +241 pp. Price £13-95. Available in the USA from Halsted Press at U.S. $24-95.

© The Palaeontological Association, 1994

Palaeontology

VOLUME 37 -PART 1

CONTENTS

Burlingiids : small proparian Cambrian trilobites of enigmatic
origin

H. B. WHITTINGTON 1

Tiny plagiaulacoid multituberculate mammals from the Purbeck
Limestone Formation of Dorset, England

Z. KIELAN-JAWOROWSKA and P. C. ENSOM 17

A new anguimorph lizard from the Jurassic and Lower
Cretaceous of England

S. E. EVANS 33

The Lower Permian synapsid Glaucosaurus from Texas

S. P. MODESTO 51

Cephalaspids from the Lower Devonian of Prince of Wales
Island, Canada

D. L. DINELEY 61

The ostracod genus Krithe from the Tertiary and Quaternary of
the North Atlantic

G. P. COLES, R. C. WHATLEY and A. MOGUILEVSKY 71

Saccocoma : a benthic crinoid from the Jurassic Solnhofen
Limestone, Germany

c. v. milsom 121

Growth and disintegration of bivalve-dominated patch reefs in
the Upper Jurassic of southern England

F. T. FURSICH, T. J. PALMER and K. L. GOODYEAR 131

A new coniferous male cone from the English Wealden and a
discussion of pollination in the Cheirolepidiaecae

K. L. ALVIN, J. WATSON and R. A. SPICER 173

Eider, shelduck, and other predators, the main producers of

shell fragments in the Wadden Sea: palaeoecological

implications 1 8 1

G. c. CADEE

The Leptoceratoidinae : small heteromorph ammonites from the
Barremian

z. vasicek and j. wiedmann 203

Printed in Great Britain at the University Press, Cambridge

ISSN 0031-0239

Published by

The Palaeontological Association • London
Price £43-00

THE PALAEONTOLOGICAL ASSOCIATION

(Registered Charity No. 276369)

The Association was founded in 1957 to promote research in palaeontology and its allied sciences.

COUNCIL 1994-1995

President : Dr R. A. Fortey, Department of Palaeontology, Natural History Museum, Cromwell Rd., London SW7 5DB
Vice-Presidents: Dr M. E. Collinson, Department of Geology, Royal Holloway and Bedford New College,
University of London, Egham, Surrey, TW20 OEX

Dr J. E. Dajlingwater, School of Biological Sciences, Williamson Building, The University, Manchester Ml 3 9PL
Treasurer: Dr T. J. Palmer, Institute of Earth Studies, University of Wales, Aberystwyth, Dyfed SY23 3DB
Membership Treasurer : Dr M. J. Barker, Department of Geology, University of Portsmouth, Burnaby Road,

Portsmouth POl 3QL

Institutional Membership Treasurer: Dr J. E. Francis, Department of Earth Sciences, The University, Leeds LS2 9JJ
Secretary: Dr J. A. Crame, British Antarctic Survey, High Cross, Madingley Road, Cambridge CB3 OET
Newsletter Reporter : Dr S. Rigby, Department of Geology, University Road, Leicester LEI 7RH (co-opted)
Marketing Manager (Sales): Dr L. Cherns, Department of Geology, University of Wales College of Cardiff CF1 3YE
Marketing Manager ( Publicity ): Dr P. B. Wignall, Department of Earth Sciences, The University, Leeds LS2 9JJ
Public Relations Officer: Dr P. A. Allison, P R. I S., The University, P.O. Box 227, Whiteknights, Reading RG6 2AB
Newsletter Editor : Dr R. B. Rickards, Department of Earth Sciences, University of Cambridge, Cambridge CB2 3EQ

Editors

Dr P. D. Lane, Department of Geology, University of Keele, Keele, Staffordshire ST5 5BG
Dr P. Doyle, School of Earth Sciences, University of Greenwich, Brewhouse Lane, London El 9NU
Dr A. R. Milner, Department of Biology, Birkbeck College, Malet Street, London WC1E 7HX
Dr C. J. Cleal, Department of Botany, National Museum of Wales, Cardiff CF1 3NP
Dr R. M. Owens, Department of Geology, National Museum of Wales, Cardiff CF1 3NP
Dr B. M. Cox, British Geological Survey, Keyworth, Nottingham NG12 5GG

Other Members: Dr M. P. Smith, School of Earth Sciences, University of Birmingham, Birmingham B15 2TT
Dr M. Purnell, Department of Geology, University Road, Leicester LEI 7RH

Overseas Representatives

Argentina: Dr M. O. Mancenido, Division Paleozoologia invertebrados, Facultad de Ciencias Naturales y Museo, Paseo del
Bosque, 1900 La Plata. Australia: Dr K. J. McNamara, Western Australian Museum, Francis Street, Perth, Western
Australia 6000. Canada: Professor S. H. Williams, Department of Earth Sciences, Memorial University, St John’s,
Newfoundland A1B 3X5. China: Dr Chang Mee-mann, Institute of Vertebrate Palaeontology and Palaeoanthropology,
Academia Sinica, P.O. Box 643, Beijing. Dr Rong Jia-yu, Nanjing Institute of Geology and Palaeontology, Chi-Ming-Ssu,
Nanjing. France: Dr J.-L. Henry, Institut de Geologie, Universite de Rennes, Campus de Beaulieu, Avenue du General
Leclerc, 35042 Rennes Cedex. Iberia: Professor F. Alvarez, Departamento de Geologia, Universidad de Oviedo, Cl. Jesus
Arias de Velasco, s/n. 33005 Oviedo, Spain. Japan: Dr I. Hayami, University Museum, University of Tokyo, Hongo 7-3-1,
Tokyo. New Zealand: Dr R. A. Cooper, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt. Scandinavia: Dr R.
Bromley, Fredskovvej 4, 2840 Holte, Denmark. USA : Professor A. J. Rowell, Department of Geology, University of
Kansas, Lawrence, Kansas 66044. Professor N. M. Savage, Department of Geology, University of Oregon, Eugene, Oregon
97403. Professor M. A. Wilson, Department of Geology, College of Wooster, Wooster, Ohio 44961. Germany: Professor
F. T. Fursich, Institut fur Palaontologie, Universitat, D8700 Wurzburg, Pliecherwall 1

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1994 are:
Institutional membership . . . £85 00 (U.S. $170) Student membership . . . . £1 1 -50 (U.S. $20)

Ordinary membership . . £28 00 (U.S. $50) Retired membership .... £14 00 (U.S. $25)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr J. E.
Francis, Department of Earth Sciences, The University, Leeds LS2 9JJ. Student members are persons receiving full-time
instruction at educational institutions recognized by the Council. On first applying for membership, an application form
should be obtained from the Membership Treasurer: Dr M. J. Barker, Department of Geology, University of Portsmouth,
Burnaby Road, Portsmouth POl 3QL. Subscriptions cover one calendar year and are due each January; they should be
sent to the Membership Treasurer. All members who join for 1994 will receive Palaeontology, Volume 37, Parts 1-4.
Enquiries concerning back numbers should be directed to the Marketing Manager.

Non-members may subscribe, and also obtain back issues up to three years old, at cover price through Blackwell
Publishers, Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, UK. For older
issues contact the Marketing Manager.

US Mailing: Second class postage paid at Rahway, New Jersey. Postmaster: send address corrections to Palaeontology ,
c/o Mercury Airfreight International Ltd, 2323 EF Randolph Avenue, Avenel, NJ 07001, USA (US mailing agent).

Cover: Early evidence of extinction: in situ tooth of extinct mastodon contrasted with that of living elephant drawn by Jan
Van Rymsdyk to illustrate William Hunter’s Royal Society paper of 1768 (see Earth Sciences History, 3, 17, fig. 1).
Reproduced by permission of the Keeper of the Hunterian Books and MSS (Glasgow University Library, Hunterian
fig. 92.3).

ARCHITECTURAL CONSTRAINTS ON THE
MORPHOGENESIS OF PRISMATIC STRUCTURE IN

BIVALVIA

by TAKAO UBUKATA

Abstract. Microscopic features of the regular simple, fibrous and non-denticular composite prismatic
structures in twenty seven living and eight fossil bivalve species have been examined. Geometrical selection and
reclination of prisms caused by competition for space were recognized in many species. The regular simple and
non-denticular composite prisms are expressed as a product of aggregated spherulites, while the fibrous prisms
as needle aggregates in druses. The main factor which causes geometrical selection of prisms is irregularity of
the settling time among prisms. The density of prisms at the first stage appears to be a principal parameter
controlling the degree of geometrical selection. The ratio of accretionary rate to growth rate of prisms at the
shell edge mainly determines whether prisms become straight or reclined. Various architectural varieties of the
prismatic structure can, therefore, be explained by the geometrical selection model originally developed for
inorganic systems.

.The study of evolution of organisms in the Earth’s history necessarily depends upon the
morphology of fossilized hard tissue. In recent years, many palaeontologists have investigated the
microstructure of such hard tissues and the process of their biomineralization, and has proved
important for systematics, morphogenesis and palaeoecology. Especially in the Bivalvia, shell
microstructure has been regarded as one of the important taxonomic characters (Boggild 1930;
Taylor et al. 1969, 1973; Carter 1980a, 1991; Uozumi and Suzuki 1981; Carter and Clark 1985).
However, the variation in shell microstructure has not yet been satisfactorily explained. Until now,
research on biomineralization has concentrated on physicochemical factors such as ion movement
(Simkiss 1976; Enyikwola and Burton 1983), nucleation (Garside 1982; Greenfield et al. 1984),
chemical composition of extrapallial fluid (Crenshaw 1972; Wada and Fujinuki 1976; Misogianes
and Chasteen 1979), and function of organic matrix (Weiner and Traub 1984; Williams 1984).

The present paper focuses on the morphogenesis of bivalve biomineralization, which links the
physical chemistry (mechanisms of formation of biominerals) and morphology (microstructure of
hard tissues). Morphogenesis of the microstructure is discussed from the point of view that
geometrical features of crystals are restricted by architectural constraints. It is generally thought
that the microstructure of hard tissues is strongly controlled by phylogenetic factors (Kobayashi
1980, 1988), but it is also important to consider adaptational or architectural aspects of the
constructional morphology (Seilacher 1970, 1973). Architectural convergence or divergence of shell
microstructure may be controlled by geometrical or spatial rather than phylogenetic constraints.
This is especially likely in the formation of molluscan shells, because they are made of closely
packed aggregations of calcite or aragonite crystals. Prismatic structure is particularly suitable for
morphogenetic analysis in relation to architectural constraints because (1) it is organizationally
simple and consists of relatively large elements (prisms) which are easily observed under a
microscope, (2) its mode of formation is assumed to be strongly influenced by spatial restriction, and
(3) it is readily preserved in fossils, so that we can deal with both extant and extinct materials.

During growth of many mineral aggregates, a characteristic competition for space between
neighbouring crystallites is common in nature (Grigor'ev 1965). Grigor'ev demonstrated that the
group growth of spherulites on an uneven surface, and the growth of druses from the randomly

| Palaeontology, Vol. 37, Part 2, 1994, pp. 241-261, 3 pls.|

© The Palaeontological Association

242

PALAEONTOLOGY, VOLUME 37

oriented incipient crystals, necessarily cause geometrical selection, i.e. termination of prisms by
competition for space. Taylor et al. (1969) interpreted the morphology of prisms in bivalve shells
using this geometrical selection model of spherulite growth, and it seems an excellent explanation
for the morphogenesis of the prismatic structures from the viewpoint of spatial restriction.
However, several questions arise. Does geometrical selection occur commonly in the formation of
various prismatic layers of bivalves? What factors cause geometrical selection, and what parameters
control the competition among prisms which is considered to determine the morphology of each
prism and appearance of the microstructure?

The purpose of this paper is to evaluate the architectural constraints on the morphogenesis of the
prismatic structure in the Bivalvia. For this purpose, the author has investigated the variation and
the mode of crystal growth of the regular simple, fibrous and non-denticular composite prismatic
structures in many living and fossil species, and considered the parameters which determine the
geometrical features of prisms during successive stages of crystal growth.

NOTES ON PRISMATIC STRUCTURE IN BIVALVIA

According to Carter and Clark (1985) and Carter et al. (1991), prismatic structure consists of
parallel, adjacent structural units (first order prisms) that do not strongly interdigitate along their
mutual boundaries. Neighbouring prisms are surrounded and bounded by a sclerotized organic
matrix. Prismatic structure has been subdivided into various groups (Boggild 1930; Carter 1980a,
19806; Uozumi and Suzuki 1981 ; Carter and Clark 1985; Kobayashi 1988; Carter et al. 1991), but
in this paper only the regular simple prismatic, fibrous prismatic and non-denticular composite
prismatic structures were investigated (Text-fig. 1).

PRISMATIC STRUCTURE

Columnar prismatic structure

Regular simple
prismatic structure

Non-denticular composite
prismatic structure

Fibrous prismatic
structure

text-fig. I . Schematic block diagrams of the three types of prismatic structure used in this paper. The informal
term ‘Columnar prismatic structure' is used to unite the regular simple and the non-denticular composite

prismatic structure.

In the nomenclature of Carter et al. (1991), regular simple prismatic structure consists of
columnar calcite or aragonite prisms whose elongation axes are arranged nearly perpendicular to
the outer shell surface. Each regular simple prism is not a single crystal (Wada 1956, 1958, 1961 ;

UBUKATA: PRISMATIC STRUCTURE IN BIVALVES

243

Watabe and Wada 1956) but lacks diverging arrangements of its second-order subunits. Non-
denticular composite prismatic structure consists of columnar prisms as in the regular simple
prismatic structure, but each prism consists of second-order diverging units toward the depositional
surface. In this paper, the author uses the informal term ‘columnar prismatic structure’ which
includes both the regular simple and the non-denticular composite prismatic structures. Fibrous
prismatic structure consists of elongate calcite or aragonite prisms which are more slender than in
regular simple prismatic structure. The elongation axes of fibrous prisms are commonly reclined or
nearly parallel rather than perpendicular to the plane of the shell surface (Carter and Clark 1985).
Suzuki and Uozumi (1981) mentioned that each fibrous prism consists of a single crystal without
any particular intraprismatic microstructure.

Prisms are initially secreted on the inner surface of the periostracum at the growing shell edge,
forming polygonal blocks on the outer shell surface. Growth of prisms is limited to one direction,
namely inward, and as prisms elongate the prismatic layer becomes thicker. The periostracum and
organic matrix are considered to play an important role in the calcification of prisms (Taylor and
Kennedy 1969; Saleuddin and Petit 1983; Wilbur and Saleuddin 1983; Lowenstam and Weiner
1989; Simkiss and Wilbur 1989). There are irregularly spaced horizontal bands of organic material
in the prisms (Watabe and Wada 1956; Wada 1957; Gregoire 1972), but it is uncertain whether or
not these horizontal organic layers interrupt crystal growth.

MATERIALS AND METHODS

A total of thirty five species of extant and fossil bivalves were studied (Tables 1-2). Their surfaces
and sections were observed by scanning electron microscopy (Hitachi S-2400) and optical
microscopy (Olympus AHBT).

Preparation for SEM

For observation of fractured surfaces, the shell in all species was broken by hand normal to the outer
shell surface and dried in air. Specimens of seven species were embedded in gypsum or epoxy resin
and polished with carborundum and diamond paste. For observation of the intraprismatic
microstructure, the polished surface was etched with 0-5 M EDTA solution buffered at pH 7-8 for
several hours, and then critical-point-dried using a Hitachi HCP-2 machine. In order to remove the
periostracum perfectly, the shells of Anodonta woodiana and Pinctada fucata were etched by a 3N
aqueous solution of sodium hydroxide or sodium hypochlorite for several days, and the outer shell
surface was observed. For observation of the initial stage of prism growth on the inner surface of
the periostracum, the shell edge adjacent to the periostracum in Atrina pectinata , Pinctada
margaritifera , Isognomon ephippiutn , I. legumen , Ostrea denselaniel/osa , A. woodiana, Mytilus
californianus and M. galloprovincialis was carefully removed from the shell under a binocular
microscope. In M. californianus and M. galloprovincialis the periostracum was etched with a
solution of sodium hypochlorite. For SEM observations, all specimens were coated with platinum
vanadium by an ion coater (Eiko IB-5).

Preparation for optical microscopy

Shells of most species were cut normal to the outer shell surface. These specimens were embedded
in gypsum or epoxy resin and polished with carborundum. The polished plane of each specimen was
etched with 5 per cent acetic acid for one minute, washed, and air dried. Subsequently an acetate
peel was prepared for each specimen (Kennish el al. 1980), which were observed under ordinary
light microscopy. Those of the prismatic shell were photographed and the number of prisms in the
scale equivalent to a 1 mm length was counted on the photograph, to measure their size and density.

244

PALAEONTOLOGY, VOLUME 37

Table I. List of extant specimens examined.

Family

Species

Sampling locality

Pteriidae

Pinctada fucata (Gould, 1850)

P. margaritifera (Linnaeus, 1758)
P. maxima (Jameson. 1901)
Pteria brevialata (Dunker, 1872)

Aburatsubo Cove, Miura Peninsula,
Kanto, Japan

Ishigaki Island, Okinawa, Japan
Indian Sea

Shikanoshima near Fukuoka, Kyushu,
Japan

Isognomonidae

Isognomon nucleus (Lamarck, 1819)
/. ephippium (Linnaeus, 1758)

I. legmen (Gmelin, 1791)

I. perna (Linnaeus, 1758)

Iriomote Island, Okinawa, Japan
Iriomote Island, Okinawa, Japan
Iriomote Island, Okinawa, Japan
Iriomote Island, Okinawa, Japan

Malleidae

Malleus regula (Forskal, 1775)

Iriomote Island, Okinawa, Japan

Pinnidae

Atrina pectinata (Linnaeus, 1758)
A. teramachii Habe, 1953
A. vexirum (Born, 1780)

Pinna muricata Linnaeus, 1758
P. nobilis Linnaeus, 1758

Osaka Bay, Kinki, Japan
Amakusa Islands, Kyushu, Japan
Singapore

Iriomote Island, Okinawa, Japan
Amakusa Islands, Kyushu, Japan

Ostreidae

Crassostrea gigas (Thunberg, 1793)
Ostrea denselamellosa Lischke, 1869

Aburatsubo Cove, Miura Peninsula,
Kanto, Japan

Iriomote Island, Okinawa, Japan

Gryphaeidae

Neopycnodonte musashiana (Yokoyama,
1920)

Amakusa Islands, Kyushu, Japan

Mytilidae

Mytilus californianus Conrad, 1837
M. galloprovincialis Lamarck, 1819

M. gray anus, Dunker, 1853

Neah Bay, California, USA
Shikanoshima near Fukuoka, Kyushu,
Japan

Soya near Wakkanai, Hokkaido, Japan

Propeamussiidae

Propeamussium sibogai (Dautzenberg and
Bavay, 1904)

P. sp.

Port Darwin, Northern Territory,
Australia

Owase, Kinki, Japan

LI nionidae

Anodonta woodiana L,ea, 1834
Cristaria plicata (Leach, 1815)

Lanceolaria oxyrhyncha (Martens, 1861)
Unio dag/asiae (Griffith and Pidgeon, 1834)

Biwa Lake, Kinki, Japan
Biwa Lake, Kinki, Japan
Biwa Lake, Kinki, Japan
Biwa Lake, Kinki. Japan

Trigoniidae

Neotrigonia margaritacea (Lamarck, 1804)

French Island, Victoria, Australia

RESULTS

Columnar prismatic structure

SEM observations were made on the fractured surface of the columnar prismatic layer in all species
listed in Tables 1 and 2. Various patterns were recognized in the geometry of the prisms. For
example, the prismatic shells of Pteria brevialata , Atrina pectinate t, A. teramachii and Neopycnodonte
musashiana are constructed of prisms approximately uniform in width (simple type; PI. 1, fig. 1). In
Isognomon perna , Malleus regula , Atrina sp. and inoceramids, however, the prisms tend to grow
continuously, showing a remarkable change in width (branched type; PI. 1, figs 2-3) (Tsujii et al.
1958; Reddy et al. 1971). In the latter case, some prisms become larger and continue to grow, but
others become smaller with growth and finally disappear within the prismatic layers, where a
peculiar feature termed ‘prism branching’ (Boggild 1930) or ‘geometrical selection’ (Taylor et al.
1969) is observed.

UBUKATA: PRISMATIC STRUCTURE IN BIVALVES

245

Table 2. List of fossil specimens examined.

Family

Species

Horizon

Locality

Inoceramidae

Inoceramus amakusensis

Uf Member, Upper Yezo

Tappu area, Rumoi

Nagao and Matsumoto,

Group; Santonian,

Province, Hokkaido, Japan

1940

Cretaceous

/. cordiformis Sowerby, 1823

Haborogawa Formation,

Haboro area, Rumoi

Upper Yezo Group;
Coniacian, cretaceous

Province, Hokkaido, Japan

/. hobetsensis Nagao and

Saku Formation, Middle

Tappu area, Rumoi

Matsumoto, 1939

Yezo Group; Turonian,
Cretaceous

Province, Hokkaido, Japan

I. undulatoplicatus Roemer,

Uij Member, Upper Yezo

Tappu area, Rumoi

1849

Group; Santonian,
Cretaceous

Province, Hokkaido, Japan

/. uwajimensis Yehara, 1924

Ubc Member, Upper Yezo

Tappu area, Rumoi

Group; Coniacian,
Cretaceous

Province, Hokkaido, Japan

Sphenoceramus naumanni

Uf Member, Upper Yezo

Tappu area, Rumoi

(Yokoyama, 1890)

Group; Santonian,
Cretaceous

Province, Hokkaido, Japan

S. schmidti (Michael, 1899)

Osoushunai Formation,

Saku area, Teshio Province,

Upper Yezo Group;
Campanian, Cretaceous

Hokkaido, Japan

Pinnidae

Atrina sp.

Hiraiga Formation, Miyako

Hiraiga area, Miyako

Group; Aptian, Cretaceous

Province, Tohoku, Japan

The elongation axes of prisms are not always arranged strictly perpendicular to the outer shell
surface; for instance, in species such as Atrina pectinata , Pinctada f neat a and Isognomon ephippium
they are, in Unio dag/asiae , Crassostrea gigas , Ostrea denselamellosa and Malleus regula they are
markedly curved and inclined toward the ventral margin (PI. 1, fig. 8). The elongation axes of prisms
and the growth line of the shell always intersect at 90°.

Geometrical selection is also seen in acetate peels of the shell in several species. For example, in
Pinctada margaritifera the density of prisms on the inner shell surface is much smaller than that on
the outer surface of the prismatic layer (PI. 1, fig. 6). Increase of average prism size inwards from
the outer shell surface was reported in Crassostrea virginica by Tsujii et al. (1958). As shown in Text-
figure 2, in all species examined the number of prisms per millimetre tends to decrease as the prisms
grow. Termination of many prisms just below the outer shell surface is clearly seen in the
honeycomb-like framework of the intercrystalline matrix in P. margaritifera (PI. 1, fig. 7). Both
simple type and branched type may represent end members of the wide variation of columnar
prismatic structure, as confirmed in Isognomon nucleus , Atrina vexirum , Pinna muricata , P. nobilis ,
Anodonta woodiana, Cristaria plicata , Union dag/asiae and Lanceolaria oxyrhyncha (PI. 1, fig. 4).
Such a remarkable variation in the size, shape and density of prisms is also observed at different
positions within a single shell. For instance, Pinctada fucata , P. margaritifera , P. maxima ,
Isognomon ephippium and I. legumen often abandon secretion of the shell at their venter, and newly
secreted prisms cover the inner surface of the abandoned shell. The size of the newly secreted prisms
is generally finer than that of the abandoned venter (PI. 1, fig. 5). Geometrical selection seems to
be prominent in such areas.

Taylor et al. (1969) pointed out that if spherulites start to grow on an uneven substratum, the
growth of some prisms will be obstructed by the lack of space (Text-fig. 3a-c). This means that each
prism may become broader if competition among prisms for occupancy of space is weak (free

246

PALAEONTOLOGY, VOLUME 37

growth of prisms). Thus microscopic observations of the initial stages of crystal growth at the
periostracal edge and of intraprismatic microstructures using living bivalves are necessary to
understand the growth mode of prisms and its controlling factors. In Pinctada margaritifera ,
Isognomon epliippium, I. legumen, Atrina pectinata , Ostrea denselamellosa and Anodonta woodiana,
many small, hemispherical prisms appear on the inner surface of periostracum at the shell edge, and
they represent an initial stage of prism formation. Subsequently, they continued to grow until they
came into contact with other prisms at their lateral sides (PI. 2, figs 1-3; Wada 1961, fig. 8). On the
outer shell surface of A. woodiana , the boundary of prisms appears to be similar to the contact of
growing circles. The boundary between two contiguous prisms is gently curved and the smaller
prism projects into the larger prism (PI. 2, fig. 4). This means that the two prisms were at different
stages of growth when they come into contact. In P. margaritifera and A. woodiana , hemispherical
prisms at the shell edge show a marked variation in their size, suggesting a difference in the
nucleation time of crystal growth among prisms (PI. 2, figs 1-2). Sides of the larger hemispherical
prisms generally curved inward, while in the smaller ones they are frequently projected toward the
outer side. Preformed organic membranes, which may inhibit the crystal growth, do not occur in
the initial stage of prism formation at the periostracal edge in any species examined.

Intraprismatic microstructure is observable in the slightly decalcified shells of Pinctada
margaritifera , Malleus regula , Atrina pectinata , Propeamussium sibogai , Anodonta woodiana , Unio
daglasiae and Neotrigonia margaritacea. From SEM observations, three types were recognized in
the intraprismatic microstructure of these species. The first, called 'spongy type’, is in the regular
simple prisms of P. margaritifera , M. regula and A. pectinata. The intraprismatic matrix of this type
appears to consist of many tablets which are nearly parallel to the shell surface (PI. 2, fig. 5). At a
higher magnification, intraprismatic organic sheets of this type exhibit a sponge-like structure (PI. 2,
fig. 6). Similar intraprismatic microstructures were reported in Pinctada fucata by Suzuki and
Uozumi (1981). Within a prism of A. pectinata , organic substances arranged parallel to the
elongation axis of the prism can be observed under the SEM (PI. 2, fig. 5). The second type, observed
in the non-denticular composite prisms of A. woodiana , U. daglasiae and N. margaritacea , appears
to consist of radially diverging aciculate elements and horizontal bands, the latter being gently
convex toward the inner direction (aciculate type; PI. 2, fig. 7). This has been described by Suzuki
and Uozumi (1981), and Carter and Lutz (1991). The third type is a foliate intraprismatic
microstructure present in propeamussiids (microfoliate type) (PI. 2, fig. 8; PI. 3, fig. 1), which was
described by Hayami (1988«, 19886) in Propeamussium watsoni and Cyclopecten bistriatus. The
foliation in this type seems to be nearly radial.

EXPLANATION OF PLATE 1

Figs 1-5. SEM photographs of vertical section of the columnar prismatic shell. 1, Atrina pectinata (Linne).
UMUT RM 19612; Recent; Osaka Bay, Kinki, Japan; simple type of the regular simple prisms; x 70.
2, Malleus regula (Forskal). UMUT RM 19613; Recent; Iriomote Island, Okinawa, Japan; branched type of
the regular simple prisms; x 35. 3, Inocerannts hobetsensis Nagao and Matsumoto. UMUT MM 19614;
Saku Formation, Middle Yezo Group, Cretaceous (Turonian); Tappu area, Rumoi Province, Hokkaido,
Japan; branched type of simple prisms; x 40. 4, Pinna nobilis Linne. UMUT RM 19615; Recent; Amakusa
Islands, Kyushu, Japan; intermediate type between simple and branched type; x45. 5, Pinctada
margaritifera (Linne). UMUT RM 19616; Recent; Ishigaki Island, Okinawa, Japan; newly secreted prisms
under an abandoned venter; x 120.

Fig. 6. Pinctada margaritifera (Linne). UMUT RM 19617; Recent; Ishigaki Island, Okinawa, Japan; optical
micrograph of acetate peel of vertical section of the regular simple prismatic shell; x40.

Figs 7-8. SEM photographs of decalcified columnar prisms. 7, Pinctada margaritifera (Linne). LIMUT RM
19618; Recent; Ishigaki Island, Okinawa, Japan; conchiolin framework of the simple prisms; x 130. 8, Unio
daglasiae (Griffith and Pidgeon). UMUT RM 19619; Recent; Biwa Lake, Kinki, Japan; reclined type of
composite prisms; x 130. In each figure, the outer surface of the shell is at the top and the ventral side to
the right.

PLATE 1

UBUKATA, bivalve prismatic structure

248

PALAEONTOLOGY, VOLUME 37

Atrina pectinata

®

n

E

3

C

T3

®

N

CO

w

c

co

w

% Isognomon ephippium

% Isognomon perna

% Inoceramus hobetsensis

Length of growing prisms

text-fig. 2. Survival rate of prisms through prism growth in six selected bivalve species. Length of growing
prisms represents the distance from the outer shell surface of the prismatic layer. Standardized number
of prisms is expressed by the ratio of the number of prisms per millimetre at different stages to the number of
prisms on the outer shell surface. Number of prisms per millimetre was counted at various positions on a single

shell in vertical section.

In Propeamussium sp., the radially elongate simple prisms and the regular simple prisms seem to
alternate throughout the growth of the prismatic layer (PI. 3, fig. 2). Among the radially elongate
simple prisms in P. sp. geometrical selection occurs in the radial direction. The regular simple prisms
in Propeamussium sibogai and P. sp. do not show geometrical selection because of their growth
mode or smaller dimensions. The growth mode of prisms in Propeamussium is still unknown,
because incipient prisms could not be found in any specimen examined.

Fibrous prismatic structure.

Fractured surfaces of the fibrous prismatic layers were examined in Mytilus calif or nianus,
M. galloprovincialis and M. grayanus. Long and narrow prisms are regularly arranged with their
elongation axes reclined or nearly parallel to the outer shell surface. However, in the early stage the
prisms are rather fine (about 0 5 pm in diameter) and exhibit an irregular spherulitic structure (PI. 3,
figs 3-4). The prisms become coarser as crystals grow and finally attain a uniform diameter
(2-5 pm) and are inclined to the shell surface. Similar changes of the size and orientation of prisms
during their growth were described by Uozumi and Suzuki (1981) in M. galloprovincialis and by
Uozumi and Iwata (1969«, 1969fi) in Mytilus coruscus. The remarkable lateral expansion observed
in the regular simple prisms of inoceramids does not occur in the fibrous prisms of the three Mytilus
species.

UBUKATA: PRISMATIC STRUCTURE IN BIVALVES

249

A

B

D

E

F

text-fig. 3. Schematic lateral views showing geometrical selection during crystal growth (after Grigor’ev 1 965).
First settlement and growth rate are regular among prisms, a-c, stages of group growth of spherulites on an
uneven substratum; a, stage of secretion of discrete spherulites; b, stage during which neighbouring spherulites
come in contact with one another and, as a result, geometrical selection occurs; c, stage of mature spherulites
showing the columnar prismatic structure, d-f, stages of group growth of single crystals; d, stage of secretion
of minute separate crystals; e, stage of druse growth showing geometrical selection; f, stage of parallel-

arranged growth.

On the shell surface of the periostracal edge, nucleation sites of prisms are localized rather than
distributed randomly or uniformly (PI. 3, fig. 6) In Mytilus californianus and M. galloprovincialis
incipient prisms on the inner surface at the periostracal edge consist of euhedral calcite crystals (PI. 3,
fig. 7), which are randomly oriented and clustered (PI. 3, figs 6-7). Until a cluster comes in contact
with surrounding clusters, crystals continue to grow spherically, but they are not typical spherulites
such as observed in turtle eggshell (Silyn-Roberts and Sharp 1986). Immediately after the formation
of clusters of prisms, the elongation axes of the prisms become irregularly oriented (PI. 3, figs 3-4).
Afterwards, each prism assumes a preferred orientation as crystal growth proceeds, and finally the
perfectly regular pattern of typical fibrous prismatic structure is established (PI. 3, fig. 5). Uozumi
and Suzuki (1981) described this as an example of geometrical selection.

In Mytilus grayemus , prisms are more or less reclined to the shell surface in the juvenile stage of
the shell, as in the case of Mytilus galloprovincialis. In the adult stage of the M. grayanus shell,
however, prisms become more strongly reclined and finally become nearly parallel to the shell
surface. At any point of the shell in the three species of Mytilus examined, the elongation axes of

250

PALAEONTOLOGY, VOLUME 37

prisms are inclined to the plane of the shell surface; in other words, prisms are extended in the radial
direction. Fibrous prisms and growth lines generally intersect at an oblique angle.

DISCUSSION

Morphogenesis of columnar prismatic structure

Grigor'ev (1965) discussed the genesis of mineral aggregates, especially the growth of druses from
randomly oriented crystals and group growth of spherulites (Text-fig. 3). According to him, the
simultaneous group growth of the spherulites on an uneven surface (Text-fig. 3a-c) and the similar
growth of various spherulites at irregular rates on an even surface (Text-fig. 4b), both necessarily
cause a geometrical selection, exemplified by malachite. Taylor et al. (1969) applied this geometrical
selection model to the growth of spherulites on an uneven molluscan shell surface to explain
morphogenesis of columnar prisms (Text-fig. 3a-c).

In this study the author confirmed that the non-denticular composite prisms of Anodonta
woodiana (PI. 2, fig. 7), Unio dag/asiae (PI. 1, fig. 8) and Neotrigonia margaritacea consist of radially
diverging aciculate subunits. This indicates that the prisms in these species are made of fine
spherulitic crystals (Taylor et al. 1969; Wilbur and Saleuddin 1983), whose morphological features
are somewhat similar to those in turtle eggshell (Silyn-Roberts and Sharp 1986). In the regular
simple prisms of Atrina pectinata , radial elements are observed, but the intraprismatic
microstructure remains uncertain, since the spherulitic microstructure is not preserved clearly in the
regular simple prisms of the present material (PI. 2, fig. 5). In their study of the early prism
formation in Pinctada radiata , Nakahara and Bevelander (1971 ) suggested that minute crystals arise
in envelopes through the prism chamber, without showing spherulitic crystal growth in the regular
simple prisms.

In the species examined with the non-denticular composite and regular simple prisms, the initial
prisms consist of hemispherical mineral aggregates (PI. 2, figs 1-3), which become modified in the
subsequent stage as a result of competition for space independent of their intraprismatic
microstructures. As for Propeamussium with microfoliated regular simple prisms, incipient prisms
and growth process of prisms were not observed. Bevelander and Nakahara (1969) proposed a
hypothetical model in which an extracellular framework of organic material is formed in the
nacreous layer before crystal formation, and that the growth of crystals is inhibited by these organic
walls (compartment model; Bevelander and Nakahara 1969, 1980; Nakahara 1979; Lowenstam
and Weiner 1989). However, the present author could not observe such a preformed organic
framework in the early stage of prism mineralization of any of the specimens examined (PI. 2, figs
1-3). This indicates that the observed morphologies are well explained by geometrical selection.
Although each prism may not always develop hemispherically throughout its growth, it should be

EXPLANATION OF PLATE 2

Figs 1-3. Incipient columnar prisms at the periostracal edge; the upper is the ventral side. 1, Anodonta
woodiana Lea. UMUT RM 19620; Recent; Biwa Lake, Kinki, Japan; x 400. 2, Pinctada margaritifera
(Linne). UMUT RM 19621 ; Recent; Ishigaki Island, Okinawa, Japan; x 200. 3, Atrina pectinata (Linne).
UMUT RM 19622; Recent; Osaka Bay, Kinki, Japan; x 650.

Fig. 4. Anodonta woodiana Lea. UMUT RM 19623; Recent; Biwa Lake, Kinki, Japan; the outer surface of
the non-denticular composite prismatic shell etched with NaOH solution; x 500.

Figs 5-8. Intraprismatic microstructure of slightly decalcified columnar prisms; the upper is the outer side.
5, Atrina pectinata (Linne). UMUT RM 19624; Recent; Osaka Bay, Kinki, Japan; spongy type; x 500. 6, as
Fig. 5; x 1300. 7, Anodonta woodiana Lea. UMUT RM 19625; Recent; Biwa Lake, Kinki, Japan; aciculate
type; x 650. 8, Propeamussium sibogai (Dantzenberg and Bavay). UMUT RM 19626; Recent; Port Darwin,
Northern Territory, Australia; microfoliate type; x 1000.

PLATE

UBUKATA, bivalve prismatic structure

252

PALAEONTOLOGY, VOLUME 37

emphasized that geometrical selection occurs a priori when prisms possibly grow laterally, and that
the appearance of a given structural type (e.g. branched type) can not be explained without lateral
growth of prisms, regardless of intraprismatic microstructure.

From the architectural viewpoint, two hierarchies can be considered in the morphogenesis of the
non-denticular composite prisms; the prism level and the intraprismatic crystal level. The growth
mode of prisms (aggregation of minute crystals) may possibly control geometrical selection,
independent of the growth mode of the intraprismatic minute crystals. Therefore, the columnar
prisms in many bivalves, excluding propeamussiids, are regarded as a product of group growth of
spherulites.

Geometrical selection of prisms caused by unevenness of the periostracum is known in
Margaritifera margaritifera (Taylor and Kennedy 1969, fig. 2). Such phenomena are, however, not
common in most species, and the outer surface of the prismatic layer and the inner surface of the
periostracal edge both appear rather smooth in most species examined (PI. 1, figs 1-2).
Consequently, unevenness of the inner surface of a periostracum may not be the main cause of
geometrical selection.

The following two factors may affect geometrically selected crystal growth in the columnar
prismatic structure of bivalves: (1) different settling time among prisms (Text-fig. 4a), and (2)
irregular growth rate within individual prisms (Text-fig. 4b). As already described, in Anodonta
woodiana and Pinctada margaritifera , hemispherical mounds of initial prisms scattered on the inner
surface of the periostracal edge exhibit a marked variation in their dimensions (PI. 2, figs 1-2),
suggesting the difference of timing in crystal initiation among them. If the growth rate of
neighbouring prisms is not constant, geometrical selection would occur in any growth stage of the
prisms (Text-fig. 4b). However, the density of prisms rapidly decreases after the initial stage of their
formation and seems to be nearly constant in the later stage (Text-fig. 2). This evidence suggests that
the growth rate is almost regular in each prism. Where growth rate is different among prisms,
geometrical selection is expected to occur when aggregates of prisms grow more or less spherically.
Such a spherical pattern of prisms is found in the hinge area of Inoceramus (PI. 3, fig. 8).
Geometrical selection commonly occurs in the prismatic shell of inoceramids (PI. 1, fig. 3), but is
absent in the prismatic hinge area of Inoceramus. This suggests an almost constant growth rate
among the columnar prisms in this region.

In summary, geometrical selection in the columnar prismatic layer appears to be caused mainly
by different nucleation times among prisms (Text-fig. 4a). According to this model, the degree of
geometrical selection may be controlled by the two factors : ( 1 ) the degree of irregularity of start of

EXPLANATION OF PLATE 3

Figs 1-2. SEM photographs of the inner surface of the simple prismatic layer of propeamussiids; the lower is
the ventral side. 1, Propeamussium sibogai (Dantzenberg and Bavay). UMUT RM 19627; Recent; Port
Darwin, Northern Territory, Australia; x 2000. 2, Propeamussium sp. UMUT RM 19628; Recent; Owase,
Kinki, Japan; x 200.

Figs 3-7. SEM photographs of the fibrous prisms. 3, Mytilus galloprovincialis Lamarck. UMLIT RM 19629;
Recent; Shikanoshima near Fukuoka, Kyushu, Japan; early stage of prisms showing irregular structure; the
arrow shows the vertical side; x 1300. 4, M. calif or nianus Conrad. UMUT RM 19630; Recent; Neah Bay,
California, USA; the outer surface of the shell showing spherulitic structure; the lower is the ventral side;
x 650. 5, M. grayanus Dunker. UMUT RM 19631 ; Recent; Soya near Wakkanai. Hokkaido, Japan; the
upper is the outer and the right is the ventral side; x 650. 6, M. californianus Conrad. UMUT RM 19632;
Recent; Neah Bay, California, USA; x400. 7, as Fig. 6; incipient prisms on the inner surface of the
periostracal edge; the lower is the ventral side; x 2000.

Fig. 8. Inoceramus hobetsensis Nagao and Matsumoto. UMUT MM 19633; Saku Formation, Middle Yezo
Group, Cretaceous (Turonian); Tappu area, Rumoi Province, Hokkaido, Japan; optical micrograph of an
acetate peel of vertical section in the hinge plate; the arrows show growing directions of prisms; x 30.

PLATE 3

UBUKATA, bivalve prismatic structure

254

PALAEONTOLOGY, VOLUME 37

text-fig. 4. Schematic lateral views show-
ing the successive stages of prism growth
starting from spherulites on a flat sub-
stratum. a, process of geometrical selec-
tion caused by unevenness of starting
time among prisms; at the mature stage,
prisms exhibit parallel-arranged struc-
ture. b, process of geometrical selection
caused by irregular growth rate within
individual prisms; termination or ex-
pansion of prisms occurs at any growth
stage. Stippled areas show terminated
prisms.

growth (Text-fig. 5a) and (2) the density of primary prismatic crystals if the prisms are not uniformly
distributed on the shell surface (Text-fig. 5b).

In the next step, local variation in the density of prisms was examined in each species. For this
purpose, the number of prisms on the outer shell surface was successively counted at intervals of
one millimetre in longitudinal thin section cut along the dorso-ventral axis. The density of prisms
on the outer shell surface is markedly variable in different parts of a shell. This suggests that the
density of prisms at the initial stage may be one of the controlling factors determining the variation
of the columnar prismatic structure, namely simple type or branched type (Text-fig. 5b). Atrina
pectinata, with typical simple type prisms, retains a large size and a relatively uniform density in the
prisms on the outer shell throughout ontogeny (PI. 1, fig. 1). On the other hand, in Inoceramus
hobetsensis , which has a higher density of prisms, geometrical selection is common in any part of
its shell (PI. 1, fig. 3). As already stated, the newly secreted prismatic layer on the inner surface of
the abandoned venter consists of relatively fine prisms as a product of marked geometrical selection
(PI. 1, fig. 5). In Pinctada fucata and P. margaritifera the density of prisms per unit interval is
markedly variable throughout their growth. At the point of the newly secreted prismatic layer on the
inner surface of the abandoned venter, the density of prisms is always small (PI. 1, fig. 5). As a result,
a close relationship exists between the density of prisms on the outer shell surface and the
termination rate of prisms of all species examined. The initial density of prisms is thus regarded as
the main factor which controls the degree of geometrical selection.

The timing of the initial deposition of prisms also has an important role in determining the
straightness of prisms. Irrespective of the existence of geometrical selection, prisms would be
oriented with their elongation axes perpendicular to the shell surface, if the starting time of prism
formation in a limited area was uniform or random (Text-fig. 4a). However, in all cases, prisms on

UBUKATA: PRISMATIC STRUCTURE IN BIVALVES

255

text-fig. 5. Two possible controls on the degree of geometrical selection, resulting in variation of the prismatic
structure, a, unevenness of starting time within individual prisms, b, density of primary prisms just below the
periostracum. As the start of growth among prisms is uneven, or as the density of initiation of prisms is high,
geometrical selection becomes significant. Regular and constant growth rate of prisms on the flat substratum
is assumed in both cases. Stippled areas show terminated prisms.

the dorsal side were made earlier than those on the ventral side. Therefore, in a limited area prisms
on the dorsal side generally tend to be extended to the venter, to obstruct the growth of the prisms
on the ventral side (Text-fig. 6). It is therefore concluded that the gradual retardation of settling time
of prisms in the longitudinal direction appears to provoke a remarkable reclination of those prisms.
In other words, the ratio of the accretionary rate of prisms to growth rate of prisms is considered
to determine the degree of reclination of prisms.

Morphogenesis of fibrous prismatic structure

According to Grigor’ev (1965), in the formation of druse of ore-bearing rocks, euhedral minerals
commonly grow with their elongation axes randomly oriented, and geometrical selection occurs
because of anisotropy of growth rate (Text-fig. 3d-f). It should be noted that in such a system the
mode of geometrical selection differs from the competition for space that occurs with expanding
spheres (Text-fig. 3a-c). The author considers that the former represents a process of competition
among minerals, while the latter is the process of competition among clusters of mineral aggregates,
e.g. spherulites. According to Grigor’ev (1965), crystals in the druse with a given preferred
orientation are selected by competition for space because of anisotropy of growth rate. Grigor’ev

256

PALAEONTOLOGY, VOLUME 37

text-fig. 6. Schematic diagrams showing
the growth of reclined columnar prisms.
In a limited area, the prisms that are
secreted earlier grow laterally and occupy
space, so that the prisms that are secreted
later are prevented from growing ver-
tically. Stippled areas show terminated
prisms.

(1965) subdivided the process of geometrical selection in genesis of druses into three stages: (1)
growth rate of separate crystals (Text-fig. 3d), (2) stage of druse growth (Text-fig. 3e), and (3) stage
of parallel-arranged growth (Text-fig. 3f).

In the present study, these three stages were recognized in the fibrous prismatic layer of Mytilus.
On the inner surface of the periostracal edge of M. californianus and M. galloprovincialis , randomly
oriented incipient prisms occur locally on the substratum (PI. 3, figs 6-7), and this corresponds to
the growth stage of separate crystals.

In the next stage, the prisms continue to grow spherically, which is similar to the stage of druse
growth. However, the stage of spherical growth of fibrous prisms may not show significant
geometrical selection because of the localization of nucleation sites (PI. 3, figs 3-4). This is quite in
contrast with the druse growth, during which competition for space is predominant. The process of
competition for space among fibrous prisms is somewhat different from quartz growth in the druse,
because initial prisms are aciculate at the stage of separate crystals, and because the rate of lateral
growth of prisms is much smaller than that of vertical growth (PI. 3, fig. 7; Text-fig. 7a). The initial
aggregate of the fibrous prisms, for example, is similar to that of geothite needles initiated on quartz
in a pegmatite vein.

When clusters come in contact with one another, competition ‘among clusters' occurs. This stage
is different from any stage in the genesis of druse, but corresponds to competition among
hemispheres of the simple prisms. Such appearance of spherulites is also observed in the
intraprismatic microstructure of the non-denticular composite prisms (PI. 2, fig. 7), but organic
sheets which envelope clusters of spherulites are not observed in the fibrous prismatic structure.
Initial sites of clusters of incipient fibrous prisms are distributed at intervals of about 20-50 pm (PI. 3,
figs 3-4, 6), and are closely correlated in order of magnitude with the intervals among initial sites
of the columnar prisms. The final stage represented by the regular arrangement of prisms may thus
correspond to the stage of parallel-arranged growth in druse (PI. 3, fig. 5; Text-fig. 7b).

As is common in druse growth, crystals perpendicular to the substratum are most likely to survive
(Text-fig. 3d-f). In the case of fibrous prisms, however, those on the dorsal side are formed earlier
than those on the ventral side. Consequently, newly secreted prisms are limited in their growth

UBUKATA: PRISMATIC STRUCTURE IN BIVALVES

257

text-fig. 7. Schematic diagrams showing
the formation of the fibrous prismatic
structure. The distribution of crystal
initiation points restricted distribution
on the inner surface of periostracal edge,
and crystals form clusters. Incipient
crystals are randomly but almost
spherically oriented (a) and subsequent
growth of crystals forms an aciculate
structure. As in reclined columnar
prisms, previously secreted clusters
inhibit the growth of later secreted
prisms, and consequently elongation
axes of the fibrous prisms are reclined to
the outer shell surface (b).

direction by already formed larger crystals (Uozumi and Suzuki 1981). Under such structural
constraint, prisms tend to recline toward the venter (Text-fig. 7b). In conclusion, the ratio of
accretionary rate of prisms to the growth rate of prisms at the shell margin appears to determine
the direction of prisms.

Phylogenetic and palaeontological implications

Prismatic structure is one of the most important shell microstructures for the taxonomy and
phylogeny of fossil Mollusca because of its frequent occurrence in fossilized hard tissue. It is widely
distributed not only in the Bivalvia (e.g. Pterioida, Ostreoida, Pectinoida, Solemyoida, Myoida,
Pholadomyoida, Nuculoida, Arcoida, Mytiloida, Veneroida, Unionoida, and Trigonioida), but also
in other members of almost all classes of the Mollusca. Moreover, prismatic structure is often
described as the primitive structure (Taylor 1973) because of its distribution in the Mono-
placophora, Archaeogastropoda and Nautiloida.

However, fabricational divergence or convergence may be derived from geometrical or spatial
constraints in architectural fabrics. As stated above, architectural varieties in each prismatic
structure can be regarded as fabricational divergence by different conditions of spatial restriction
in the morphogenesis of prisms. Such architectural varieties do not necessarily reflect phylogenetic
constraints, because they are observed at different positions within a single shell in Pinctada and
Isognomon (PI. 1 . fig. 5). Such fabricational divergence may be derived in other shell microstructures,
e.g. nacreous, foliated and crossed lamellar structures.

Carter and Clark (1985) suggested that the most shell microstructures are characterized by
convergence at various taxonomic levels. Particularly prismatic structure is organizationally simple
and may be easily fabricated in nature. 'Prismatic '-like structures are found in honey-comb.

258

PALAEONTOLOGY, VOLUME 37

colonial coral, epidermal cells of some plants, and even in inorganic minerals, which are clearly not
'monophyletic’. The regular simple prisms in propeamussiid species indicate the fabricational
convergence, i.e. polyphyly of regular simple prismatic structure, because intraprismatic
microstructure in propeamussiids is quite different from that in other bivalve species which have a
regular simple prismatic shell. The regular simple and non-denticular composite prisms are both
expressed as a product of a hemisphere at the first order structural level (except propeamussiids),
but intraprismatic microstructure of the regular simple prisms is distinguished from that of the non-
denticular composite prism by lack of radially diverging aciculate elements. This fact may indicate
fabricational convergence of the mode of growth of the first order structural units in columnar
prismatic structure.

In this paper, the author only introduces the potential of the morphogenetic aspect of the skeletal
microstructure for phylogeny and taxonomy and emphasizes that palaeontologists should take it
into account in any consideration of the phylogenetic significance of skeletal microstructures in
fossil organisms.

CONCLUSIONS

1. Marked variation in the size and density of crystals was observed in the columnar prismatic
structure of thirty two bivalve species belonging to ten families. For example, the simple type of this
structure is constructed by prisms approximately uniform in width, while the branched type consists
of prisms with variable width, as a result of competition for space among prisms and consequent
geometrical selection. Prisms are not always perpendicular to the outer shell surface, being
occasionally oblique to it.

2. Hemispherical incipient columnar prisms were observed at the periostracal edge in Pinctada
margaritifera , Isognomon ephippium , /. legumen , Atrina pectinata , Ostrea denselamellosa , and
Anodonta woodiana. Both regular simple calcite, and non-denticular composite aragonite, prisms
generally grow almost spherically in the species examined, except in propeamussiids.

3. It is considered that geometrical selection in the columnar prismatic structure of bivalves is
caused by irregularity in the initial time of deposition among prisms. The degree of geometrical
selection appears to be controlled mainly by the initial density of prisms on the inner surface of the
periostracum at the ventral shell edge.

4. Incipient fibrous prisms in Mytilus californianus and M. galloprovincialis seem to consist of
euhedral and needle-like calcite crystals without a preferred orientation. Nucleation sites of the
fibrous prisms in the two species appear to aggregate. Consequently prisms grow almost spherically,
forming a cluster.

5. When clusters of fibrous calcite prisms come in contact with one another, aggregations of prisms
appear to become irregular because of the dominance of geometrical selection. The mode of
geometrical selection in the process of competition for space in fibrous prisms differs from that in
columnar prisms, because the latter shows typical spherulitic growth. The mode of geometrical
selection of fibrous prisms corresponds to that of needle-like crystal aggregates in druses.

6. Reclination of the columnar and fibrous prisms is caused by the retardation of nucleation time
of prisms in the longitudinal direction and the ratio of accretionary rate versus growth rate of
prisms.

7. Architectural varieties in each prismatic structure are regarded as fabricational divergence.
Various intraprismatic microstructures of the columnar prisms indicate fabricational convergence
of columnar prismatic structure.

Acknowledgements. The author expresses his appreciation to Professor Itaru Hayami and Professor Kazushige
Tanabe (University of Tokyo) for helpful advice and continuing encouragement during the course of this work,
and for the critical reading of the first draft. The author is also grateful to Professor Joseph Carter (University
of North Carolina) for critical reading of the typescript and valuable comments. Several stimulating
discussions with Drs Katsumi Abe (Shizuoka University), Tatsuo Oji (University of Tokyo), Yasunari Shigeta
(Mikasa Museum), Mr. Shinji Isaji (University of Tokyo), and other colleagues of the Geological Institute in

UBUKATA: PRISMATIC STRUCTURE IN BIVALVES

259

University of Tokyo are also greatly acknowledged. Thanks are due to Ryukyu Pearl Inc. for providing the
living specimens of Pinctada margaritifera.

REFERENCES

bevelander, G. and nakahara, h. 1969. An electron microscope study of the formation of nacreous layer in
the shell of certain bivalve molluscs. Calcified Tissue Research 3, 84-92.

— 1980. Compartment and envelope formation in the process of biological mineralization. 19-27. In
omori, M. and watabe, n. (eds). The mechanisms of biomineralization in animals and plants: proceedings of
the third international biomineralization symposium. Tokai University Press, Tokyo, xiii + 310 pp.
boggild, o. b. 1930. The shell structure of Mollusks. Det Kongelige Danske Videnskabernes Selskabs Skrifter ,
Naturvidenskabelig og Mathematisk , Afdeling , Series 9, 2, 231-326.
born, i. von. 1780. Testacea musei caesarei vindobonensis. Imperial Museum, Vienna, 442 pp.
carter, j. G. I980fl. Environmental and biological control of bivalve shell mineralogy and microstructure.
69-113. In rhoads, D. c. and lutz, r. a. (eds). Skeletal growth of aquatic organisms. Plenum Publishing
Corporation, New York, xiii + 750 pp.

— 1980A Guide to bivalve shell microstructure. 645-673. In rhoads, d. c. and lutz, r. a. (eds). Skeletal
growth of aquatic organisms. Plenum Publishing Corporation, New York, xiii + 750 pp.

— 1991. Evolutionary significance of shell microstructure in the Palaeotaxodonta, Pteriomorphia and
Isofilibranchia (Bivalvia: Mollusca). 135-296. In carter, j. g. (ed. ). Skeletal biomineralization: pattern ,
process and evolutionary trends , Volume 1. Van Nostrand, New York, vii + 832 pp.

— and Clark, G. R. ii. 1985. Classification and phylogenetic significance of molluscan shell microstructure,
15-71. In broadleaf, t. w. (ed.). Mollusks: notes for a short course. Department of Geological Science ,
Studies in Geology , B. University of Tennessee. Knoxville, Tennessee, ix + 305 pp.

— and sixteen others. 1991. Glossary of skeletal biomineralization. 337-399. In carter, j. g. (ed.). Skeletal
biomineralization : pattern , process and evolutionary trends , Volume I . Van Nostrand, New York, vii + 832 pp.

— and lutz, r. a. 1991. Bivalvia (Mollusca). 5-28. In carter, j. g. (ed.). Skeletal biomineralization:
patterns , process and evolutionary trends , Volume 2, Atlas and index. Van Nostrand, New York, 101 pp.

conrad, t. a. 1837. Descriptions of new marine shells from upper California, collected by Thomas Nuttal.

Esquire Journal of Academy , Natural Science , 7, 227-268.

Crenshaw, M. a. 1972. The inorganic composition of molluscan extrapallial fluid. Biological Bulletin , Marine
Biological Laboratory , Woods Hole , Massachusetts , 143, 506-512.
dautzenberg, p. and bavay, a. 1904. Description d’un Ammussium nouvean drague per le Siboga daus la nrer
de Celebes. Journal de Conchy liologie , 52(3), 207-21 1.
dunker, w. 1853. Neue Mytilaceen. Zeitschrift fur Mctlakozoologie , 19, 82-92.

1872. Die Gattung Avicula. Conchylien Cabinet , 7 (3), 1-84.

enyikwola, o. and burton, r. f. 1983. Chloride-development electrical potentials across the mantle epithelium
of Helix. Comparative Physiology, 74, 161-164.

forskal, R. 1775. Descriptions Animalium Avium , Amphibiorum , Piscium , Insectorum , Vermium. Leipzig,
164 pp.

GARSIDE, j. 1982. Nucleation. 23-25. In nancollas, g. h. (ed.). Biological mineralization and demineralization.
Springer- Verlag, Berlin, 415 pp.

gmelin, L F. 1791. Systema naturae per tria vegua naturae , secundum classes , ordines , genera , species cum
characteri bus, cliff erentiis, synonymis, locis, eaitson 13. G. E. Beer, Leipzig, 3021-3910 pp.

GOULD, a. A. 1850. The shells from the exploring expedition. Proceedings of the Boston Society of Natural
History , 3, 252-258.

greenfield, E. m., wilson, d. c. and crenshaw, m. a. 1984. Inorganic nucleation of calcium carbonate by
molluscan matrix. American Zoologist, 24, 925-932.

gregoire, c. 1972. Structure of molluscan shell. 45-102. In florkin, m. and scheeler, b. t. (eds). Chemical
zoology. Volume 7. Academic Press, New York, 567 pp.

Griffith, E. and pidgeon, e. 1834. The Mollusca and Radiata. Cuvier's Animal Kingdom, Volume 12. London,

601 pp.

grigor’ev, d. p. 1965. Ontogeny of minerals. Israel Program for Scientific Translations, Jerusalem, v + 250 pp.
habe, t. 1953. Pinnidae, Placunidae and Anomidae in Japan. 1-6. In kuroda, t. (ed.). Illustrated catalogue of
Japanese shells. Volume 24. Kyoto, 6 pp.

hayami, i. 1988m Functional and taxonomic implications of internal ribs of Propeamussium. Transactions and
Proceedings of the Palaeontological Society of Japan, New Series, 150, 476^190.

260

PALAEONTOLOGY, VOLUME 37

hayami, I. 1988/). [The taxonomic character of propeamussiids from Japan], Venus , the Japanese Journal of
Malacology , 47, 71-82. [In Japanese],

jameson, H. L. 1901. On the identity and distribution of the mother of pearl oysters, with a revision of the
subgenus Margaritifera. Proceedings of the Zoological Society of London, 71, 372-394.
kennish, m. j., lutz, r. a. and rhoads, d. c. 1980. Preparation of acetate peels and fractured sections for
observation of growth patterns within the bivalve shell. 597-601. In rhoads, d. c. and lutz, r. a. (eds).
Skeletal growth of aquatic organisms. Plenum Publishing Corporation, New York, xiii + 750 pp.
kobayashi, i. 1980. Various patterns of biomineralization and its phylogenetic significances in bivalve
molluscs. 145-155. In omori, m. and watabe, n. (eds). The mechanisms of biomineralization in animals and
plants : Proceedings of the third international biomineralization symposium. Tokai University Press, Tokyo,
xiii + 310 pp.

— 1988. [Shell structure of bivalvian molluscs and its phylogenetic evolution]. 97-1 12. In omori, m., suga,
a. s. and goto, M. (eds). Biomineralization and phytogeny of marine organisms. Tokai University Press,
Tokyo, 305 pp. [In Japanese].

lamarck, j. b. p. a. de m. 1804. Sur une nouvclle espece de Trigonie, et sur une nouvelle Huitre, decouvertes
dans le voyage du capitaine Baudin. Annuaire du Museum National d Histoire Naturelle , 4, 1-355.

1819. Historic naturelle des animaux sans vertebres. 6 (1). Verdiere, Paris, 343 pp.
lea, i. 1834. Description of new freshwater and land shells. Transactions of the American Philosophical Society ,
6(1), 1-108.

leach, w. E. 1815. The zoological miscellany , being descriptions of new , or interesting animals. Volume 1.
London 150 pp.

linnaeus, C. 1758. Systema naturae per tria vegua naturae , secundum classes , or dines , genera , species cum
characteri bus , different iis , synonymis , locis , editison JO, reformata, Tomus 1 . Laurentii, Stockholm, 823 pp.
i ischke, c. E. 1869. Diagnosen neuer meeres Conchylien von Japan. Malakozoologische Blatter , 16, 105-109.
lowenstam, H. A. and weiner, s. 1989. On biomineralization. Oxford University Press, New York, ix + 324 pp.
martens, E. von. 1861. On the mollusca of Siam. Proceedings of the Zoological Society of London, 29, 6-18.
Michael, R. 1899. Uber Kreidefossilien von der Insel Sachalin. Jahrbuch der Koniglich Preussischen
Geologischen Landesanstalt und Bergakademie zu Berlin, 19, 153-164.
misogianes, m. j. and chasteen, n. d. 1979. Extrapallial fluid: a chemical and spectral characterization of the
extrapallial fluids of Mytilus edulis. Analytical Biochemistry, 100. 324-334.
nagao, T. and matsumoto, T. 1939. A monograph of the Cretaceous Inoceramus of Japan. Part 1. Contribution
from the Department of Geology and Mineralogy , Faculty of Science, Hokkaido Imperial University , 4.
241-299.

— 1940. A monograph of the Cretaceous Inoceramus of Japan. Part 2. Contribution from the Department of
Geology and Mineralogy, Faculty of Science, Hokkaido Imperial University, 6, 1-64.

nakahara, h. 1979. An electron microscope study of the growing surface of nacre in two gastropod species.
Turbo cornutus and Tegula pfeifferi. Venus, the Japanese J owned of Malacology, 47, 205-21 1.

— and bevelander, g. 1971 . The formation and growth of the prismatic layer of Pinctada radiata. Calcified
Tissue Research, 7, 31-45.

reddy, K. R., jagadiswara rao and chakrapani naidu, M. G. 1971. The shell structure of Inoceramus from
the Upper Cretaceous beds near Ariyalur, South India. Journal of the Geological Society of India, 12, 80-83.
roemer, f. 1849. Texas. Mit besonderer Riicksicht auf Deutsche Auswanderung und die physischen Verhaltnisse
des Landes. Bonn, xiv + 464 pp.

saleuddin, a. s. m. and petit, h. p. 1983. The mode of formation and the structure of the periostracum.
199-234. In saleuddin, a. s. m. and wilbur, k. m. (eds). The Mollusca, Volume 4, Physiology, Part I.
Academic Press, New York, xx + 523 pp.

seilacher, a. 1970. Arbeitskonzept zur Konstruktions-Morphologie. Lethaia, 3, 343-396.

1973. Fabricational noise in adaptive morphology. Systematic Zoology, 22, 451-465.
silyn-roberts, h. and sharp, r. m. 1986. Crystal growth and the role of the organic network in eggshell
biomineralization. Proceedings of the Royal Society of London, Series B, 227, 303-324.
simkiss, k. 1976. Cellular aspects of calcification. 1-31. In watabe, n. and wilbur, k. m. (eds). The mechanism
of biomineralization in the invertebrates and plants. University of South Carolina Press, Columbia, 461 pp.

— and wilbur, k. m. 1989. Biomineralization : cel! biology and mineral deposition. Academic Press, San
Diego, California, xiv + 337 pp.

sowerby, j. de c. 1823. The mineral conchology of Great Britain. Volume 5. London, 168 pp.

Suzuki, s. and uozumi, s. 1981. Organic components of prismatic layers in Molluscan shell. Journal of the
Faculty of Science, Hokkaido University, Series 4, 20, 7-40.

UBUKATA: PRISMATIC STRUCTURE IN BIVALVES

261

1973. The structural evolution of the bivalve shell. Palaeontology. 16, 519-534.

— and Kennedy, w. j. 1969. The influence of the periostracum on the shell structure of bivalve molluscs.
Calcified Tissue Research 3. 274-283.

— and hall, a. 1969. The shell structure and mineralogy of the Bivalvia. Introduction. Nuculacea-
Trigonacea. Bulletin of the British Museum (Natural History ), Zoology Series , Supplement , 3. I 1 24.

— 1973. The shell structure and mineralogy of the Bivalvia. II. Lucinacea-Clavegellacea,
Conclusions. Bulletin of the British Museum (Natural History ), Zoology Series , 22, 253-284.
thunbfrg, c. p. 1793. Techning och beskrifning pa eu stor Ostronsort infran Japan. Kungliga Svenska
Vetenskapsakademiens Handlingar , 14, 140-142.

tsujii, t., sharp, d. G. and wilbur, k. m. 1958. Studies on shell formation. VII. The submicroscopic structure
of the shell of the oyster Crassostrea virginica. Journal of Biophysical and Biochemical Cytology , 4. 275-280.
uozumi, s. and iwata, k. 1969a. [Studies on calcified tissues, part I. Ultrastructure of the conchiolin in Mvtilus
coruscus Gould], Earth Science ( Chikyu Kagaku ), 23, 1-6. [In Japanese].

1969b. [Studies on calcified tissues, part II. Composition of ultrastructure of the organic matrix
in prismatic region of recent and fossil Myti/us]. Journal erf the Geological Society of Japan, 75. 417-428. [In
Japanese],

- 1981. [The evolution of shell structure in the Bivalvia] 63-77. In habe, t. and omori, m. (eds). Studies
of mollusc an paleobiology; Professor Omori memorial volume. Publication Committee, Niigata University,
Niigata, 366 pp. [In Japanese].

wada, k. 1956. [Electron-microscopic observations on the shell structure of the pearl oyster (Pinctada
martensii). I. Observations on the calcite crystals in prismatic layer]. Bulletin of the National Pearl Research
Laboratory , 1, 1-6. [In Japanese].

— 1957. [Electron-microscopic observations on the shell structure of the pearl oyster (Pinctada martensii).
III. On the laminary structures of the shell]. Bulletin of the National Pearl Research Laboratory, 2, 86-93.
[In Japanese],

- 1958. [Shell structure and development in Pinctada martensii (Dunker). I. Crystalline structure]. Bulletin
of the National Peart Research Laboratory, 4, 261-274. [In Japanese].

- 1961. Crystal growth of the molluscan shell. Bulletin of the National Pearl Research Laboratory, 7,
703-828.

— and funjinuki, t. 1976. Biomineralization in bivalve molluscs with emphasis on the chemical composition
of the extrapallial fluid. 175-190. In watabe, n. and wilbur, k. m. (eds). The mechanisms of mineralization
in the invertebrates and plants. University of South Carolina Press, Columbia, 461 pp.

watabe, n. and wada, k. 1956. On the shell structure of Japanese pearl oyster, Pinctada martensii (Dunker).

(I). Prismatic layer. I. Report of the Faculty of Fisheries , Prefectural University of Mie, 2, 227-232.
weiner, s. and traub, w. 1984. Macromolecules in the mollusc shell and their functions in biomineralization.

Philosophical Transactions of the Royal Society of London, Series B, 304, 425-434.
wilbur, k. m. and saleuddin, a. s. m. 1983. Shell formation. 235-287. In saleuddin. a. s. m. and wilbur.

k. M. (eds). The Mollusca, Volume 4, Physiology, Part I. Academic Press, New York, xx + 523 pp.
williams, r. J. p. 1984. An introduction to biominerals and the role of organic molecules in their formation.

Philosophical Transactions of the Royal Society of London, Series B, 304, 41 1-424.
yehara, s. 1924. On the Izumi sandstone group in the Onogawa Basin, Prov. Bungo, and the same group in
Uwajima, Prov. Iyo. Japanese Journal of Geology and Geography, 3, 24-40.
yokoyama, m. 1890. Verstein. Japan. Kreide. Palaeontographica, 36. 159-202.

1920. Fossils from the Miura Peninsula and its immediate north. Journal of the College of Science,
Imperial University of Tokyo, 39 (6), 1-198.

TAKAO UBUKATA

Geological Institute
University of Tokyo
7-3-1, Hongo, Bunkyouku
Tokyo, Japan

Typescript received 16 April 1993
Revised typescript received 12 October 1993

DINOFLAGELLATE CYSTS FROM THE
GLACIAL/POSTGLACIAL TRANSITION IN
THE NORTHEAST ATLANTIC OCEAN

by REX HARLAND

Abstract. Higli-resolulion dinoflagellate cyst analysis of two DSDP holes and two British Geological Survey
cores, from the drift ridges on the south-western and south-eastern flanks of the Rockall Plateau and the
continental slope off western Scotland respectively, has yielded detailed cyst spectra across the glacial/
postglacial transition. These spectra illustrate clearly the substantial climatic and palaeoceanographic changes
that have accompanied deglaciation over the last 13 Ka and the enigmatic short-term return to the cooler
conditions of the Younger Dryas. The dinoflagellate cyst assemblages have undergone changes both in species
composition and in the numbers of cysts per gram recovered but show consistent and repeatable evidence of
fluctuations within the dinoflagellate cyst floras. Comparisons are made with earlier studies and with recently
published work from the Norwegian Sea, which together indicate substantial detail within the Holocene
climatic record, and are interpreted as representing considerable oceanographic variability throughout the last
10 Ka.

Over recent years, there has been increased interest in the nature and timing of the
glacial/postglacial transition in the north-eastern Atlantic Ocean. Attention to this phenomenon
has continued to grow since the advent of the CLIMAP studies (Cline and Hays 1976) and the
increased availability of DSDP and ODP material, and of the many cores taken by other research
organizations including the British Geological Survey (BGS). This interest in the last deglaciation
partly stems from an increased ability to model the dynamic changes involved in the shift from a
fully glacial world, at about 18 Ka, to the "interglacial’ conditions of today and partly in its
implications for the future.

The last deglaciation was not a gradual event but rather non-linear in character (Duplessy et al.
1986; Bard et al. 1987). In particular, the importance of a short-lived and unexpected reversal in the
warming trend, the Younger Dryas, was realized. The dynamics of this reversal and the rate of
change of the underlying causes have proved to be of great concern, and are particularly relevant
to predictive modelling of oceanographic change as the greenhouse effect vies with the orbital-
forcing mechanisms in determining the possible future climate of the Earth.

This Younger Dryas reversal, in an otherwise warming trend, fits into the currently understood
deglaciation history as follows (after Ruddiman 1987). The period from full glacial conditions
(18 Ka) to about 13 Ka included winter sea-ice to about 50 °N in the North Atlantic Ocean but
perhaps with sufficient annual variation to have large ice-free areas during some winters; sea level
at the time of maximum glaciation was about 121 m below present sea-level (Fairbanks 1989).
During the period 13 to II Ka the polar front retreated to the north-west allowing warmer, more
saline, water from the subtropical gyre into the eastern and central North Atlantic. This was then
followed from 1 1 to 10 Ka (the Younger Dryas) by an advancement of the polar front to a location
close to its position during full glacial times. After 10 Ka the polar front again retreated to the
north-west into the south-eastern Labrador Sea, where it remains today.

It has been suggested from foraminiferal evidence that there was seasonal ice cover in the
Norwegian Sea during the Younger Dryas (Sejrup et al. 1984). This was before the later influx of
the North Atlantic Current (NAC), as the polar front retreated to its present position, at about

| Palaeontology, Vol. 37, Part 2, 1994, pp. 263-283. |

© The Palaeontological Association

264

PALAEONTOLOGY, VOLUME 37

10 Ka. The Holocene warming continued from 10 Ka to a maximum interglacial configuration at
about 6 Ka (Ruddiman and McIntyre 1981). Spectral analysis of Late Pleistocene/Holocene
sediments has revealed a periodicity of 380 yrs and 2-6 Ka akin to the known 14C fluctuations in the
atmosphere associated with solar activity (Pisias et al. 1973; Keigwin and Jones 1989). Historically,
this cyclicity may have been recognized earlier with the establishment of the Blytt-Sernander
vegetational zones of continental Scandinavia (Sernander 1908). These vegetational changes also
appear to have a cyclicity of about 2-5 Ka, like that recorded by Denton and Karlen (1973) in a
study of the Holocene glacial extensions in the Yukon and Alaska.

Furthermore, the Younger Dryas is also remarkable for the rate of change that marks its
inception and end. The advance of the polar front at the beginning of the period appears to have
occurred over less than 100 years (Ruddiman 1987) and its retreat in as little as 20 years (Dansgaard
et al. 1989). The assessment of these rates of change is dependent on good chronostratigraphical
control. Recent dating of a core by Bard et al. (1987), taken in the Rockall area close to the present
study area, placed the Younger Dryas between 11500 and 10700 YBP. The Younger Dryas is,
therefore, commonly regarded as being about 1000 years in duration. This is too short for a simple
response to orbital-forcing (Ruddiman 1987), but certainly not too short a time for the North
Atlantic Ocean to undergo major changes in its circulation regime. The origin and timing of this
‘catastrophe’ is not understood, but may involve a small disturbance unbalancing a somewhat
chaotic system (Berger 1990).

Many of the explanations currently espoused involve the shutting down of the production of
North Atlantic Deep Water (NADW) by, for instance, the lowering of the North Atlantic surface
temperature and reducing the flux of nutrient-depleted northern source water into the deep Atlantic
(Boyle and Keigwin 1987). One way of lowering the surface temperature is by introducing cold,
fresh meltwater from the waning Laurentide ice sheet. This would also have the effect of reducing
surface salinities and density, and so lessening the efficacy of the formation of NADW, thus slowing
the North Atlantic ‘conveyor’ (Broecker et al. 1988, 1989). Recently it has been suggested that the
ice sheet disintegration occurred in two steps, as evidenced by rapid rises in sea-level at 12 and
9-5 Ka (Fairbanks 1989) and by the accompanying major decreases in lsO and increased deep water
formation and ventilation during the Younger Dryas (Jansen and Veum 1990). Although the
Younger Dryas is regarded as a brief return to glacial conditions, Jansen and Veum (1990) believed
that the formation of NADW in the Younger Dryas is more like the situation today than that at
full glacial times. Finally, the recent work of Kudrass et al. (1991 ) and Mathewes et al. ( 1993) belies
the notion of the Younger Dryas being a predominantly north-west European phenomenon. This
argues against a parochial North Atlantic cause and may point to global lowering of atmospheric
CO, (Kudrass et al. 1991).

Undoubtedly the controversy will continue, and remain inextricably linked to the production and
strength of NADW. Recent insights suggest both a rapid shutdown of NADW circulation during
the Younger Dryas, based upon geochemical data derived from planktonic foraminifera (Lehman
and Keigwin 1992), and that deep ventilation was as vigorous then as it is today, based upon benthic
foraminiferal isotope information (Veum et al. 1992). The timing and nature of the last deglaciation
is a complex phenomenon, but vitally important to the testing of our predictive skills and perhaps
finally to our future well-being.

The ability to model and interpret this last deglaciation requires not only an understanding of the
physical changes, but also the consequences of these changes on the biological component. The
effects of deglaciation have long been demonstrated by both floral and faunal populations, but it
is only recently that dinoflagellate cysts have been utilized to provide information for both climatic
and oceanographic reconstruction (Harland 1988; De Vernal et al. 1992). This has proved possible
only in the light of increasing knowledge of the ecology of dinoflagellates and their cysts ( Dale 1 983 ;
Harland 1988). Recently, quantitative attempts have been made to use dinoflagellate cyst
assemblages as proxies for palaeotemperature estimates (Edwards et al. 1991, Mudie 1992).

Dinoflagellates are part of the plankton, but their cysts act as benthos (Dale 1983) providing a
unique opportunity to access information from both these important regimes within the ocean.

HARLAND: NORTH ATLANTIC DINOFLAGELL ATES

265

However, careful consideration of this information is required before any interpretations are
attempted (Evitt 1985) since the cyst assemblages do not exactly mirror the motile dinoflagellate
populations. A commentary on dinoflagellates, their cysts and their usefulness in elucidating
Quaternary climatic change may be found in Harland (1988, 1 992<r/).

It is against this background that the dinoflagellate cyst record of the Glacial/Postglacial
transition is being studied to assess its contribution to the understanding of the last deglaciation of
the North Atlantic Ocean. Turon (1978, 1980, 1981) has demonstrated that indeed there were
significant changes in the dinoflagellate cyst assemblages during this time and that they may relate
to movements of the surface water masses. Recent investigations (Stoker et al. 1989) have indicated
the potential of dinoflagellate cyst analysis in elucidating the effects of bottom water currents,
whereas most recently Baumann and Matthiessen (1992) have used dinoflagellate cyst assemblages
and coccolith data to investigate surface water mass conditions through the Holocene of the
Norwegian Sea. This contribution aims to further the use of dinoflagellate cysts in assisting the
understanding of the last deglaciation and the postglacial history of the northeastern Atlantic
Ocean.

MATERIALS AND METHODS

Four cored sequences that are known to span the Glacial/Postglacial transition were chosen for
study. Their localities and water depths are indicated in Table 1 and Text-figure 1 . Sample numbers
and depth/core data are given in Appendix 1 ; all samples, preparations and data sheets are held in
the palynological collections of the Biostratigraphy and Sedimentology Group, BGS, Keyworth.

table I Core Localities.

Core

Location

Water depth (m)

Samples

DSDP 552A

Lat: 56° 02-56' N
Long: 23° 13-88' W

2301

10

DSDP 610A

Lat: 53° 13-30' N
Long: 18° 53-21' W

2417

10

BGS Vibrocore 57/-10/84

Lat: 57° 37-33' N
Long: 09° 49-19' W

1346

21

BGS Gravity Core 57/-10/47

Lat: 56° 23-22' N
Long: 09° 57-98' W

1787

21

Standard preparation techniques were used throughout but the numbers of cysts per gram of
sediment were always calculated using the method described in Harland ( 1989). The majority of the
subsequent text-figures, using these data, are drawn to the same scale to allow close comparison;
however, the scale for Text-figure 4 has been adjusted to allow for formatting. Some measure of
chronostratigraphical control is provided by the oxygen isotope stratigraphy where available. The
two sets of data from the DSDP Holes are aliquots from samples used to establish the stable isotope
stratigraphy and, therefore, offer a first-order correlation between the biostratigraphy and the stable
isotope data. The oxygen isotope data for DSDP Hole 552A is based upon that quoted for the
benthic forammifer Cibicidoides wuellerstorfi (Schwager) whereas that from DSDP Hole 61 0A is
based upon the planktic foraminifer Neogloboquadrina pachyderma (Ehrenberg). All chronological
control for the Younger Dryas is based upon the work of Bard et al. (1987). In addition a taxonomic
listing of the taxa encountered is provided in Appendix 2 with reference to recent illustrations.
Discussions on the taxonomy of Quaternary dinoflagellate cysts are to be found in Harland ( 1982,
1983).

266

PALAEONTOLOGY, VOLUME 37

Norwegian Sea Overflow Water (NSOW)

North Atlantic Deep Water (NADW)
Antarctic Bottom Water (ABW)

text-fig. 1. Location map for DSDP Sites 552 and 610 together with BGS Sites 56/-10/47 and 57/-10/84.
Contourite drift deposits are stippled and the various bottom water currents are shown.

HARLAND: NORTH ATLANTIC DINOFLAGELLATES

267

RESULTS

Each of the four cores will be discussed in turn before more general interpretations and conclusions
are reached and comparisons made. Unfortunately, high-resolution dinoflagellate cyst records are
rare for this area (De Vernal el al. 1992), highlighting the relative paucity of dinoflagellate cyst
analyses in investigating Late Pleistocene and Holocene sequences.

DSDP Hole 552A

Originally drilled as a part of DSDP Leg 81 on the south-western margin of the Rockall Plateau,
this core proved a succession of sediments on the Hatton Drift (Roberts et al. 1984). It was one of
the first DSDP holes to be drilled using the Hydraulic Piston Corer (HPC) and it provided a detailed
stable isotope record indicating the initiation of glacial conditions at about 2-37 Ma in the North
Atlantic, some 700 Ka before the Pliocene/Pleistocene boundary (Shackleton and Hall 1984). Also
it provided evidence to suggest that, between 0 9 and 0-7 Ma, climatic cycles responding to the
orbital obliquity rhythm (41 Ka) changed to ones responding to the eccentricity rhythm (100 Ka)
(see Ruddiman and Wright 1987). The oxygen isotope signal from Hole 552A for the time span of
interest, i.e. the youngest 1 m, is reproduced in Text-figure 2 after Shackleton and Hall (1984). The
record appears to show a reasonable Termination I A, a poorer Termination IB and the Younger
Dryas Interstadial ; this allows for some chronostratigraphical control and the deduction of a
maximum sedimentation rate of 300 mm per Ka during the Younger Dryas and about 70 mm per
Ka for the Holocene. As discussed later, however, it is more than likely that the younger part of the
Holocene is missing. Morton (1984) reported the presence of volcanic ash within Core 1 between
samples at 0 38 m and 0-68 m with an extrapolated peak at 0-50 m; he tentatively correlated it with
North Atlantic ash level I reported by Ruddiman and Glover (1975), dated at 9400 YBP, but now
regarded as 10600 YBP (see Kvannne el al. 1989); this is also included on Text-figure 2.

The dinoflagellate cyst record of the Quaternary of Hole 552A was first discussed by Harland
(1984) and subsequently by Harland (1989, 19926) but detail of the last deglaciation was never
described. Originally termed ‘event 3’ (Harland 1984) from Core 1, Section 1, it marked the
incoming of rich and diverse dinoflagellate cyst assemblages and a changeover of dominance from
Bitectatodinium tepikiense Wilson to Operculodinium centrocarpum (Deflandre and Cookson) Wall.
This was noted as marking the onset of the present ameliorative climate. More detail of the
dinoflagellate cyst assemblages of isotope stage 1 was given in Harland (1989) and interpreted in
terms of the biological oceanography. However, until now no discussion of the deglaciation history
has been given.

It is clear from Text-figure 2 and the data documented in Table 2 that the earlier part of the
record, assumed to be Late Glacial in age and certainly > 10700 YBP, is characterized by a single
productive sample at 0-9 m. This sample contains a dinoflagellate cyst assemblage characterized by
low numbers of cysts per gram of sediment (< 50) and the species Bitectatodinium tepikiense ,
Nematosphaeropsis labyrinthus (Ostenfeld) Reid, Operculodinium centrocarpum and Protoperidinium
pentagonum (Gran) Balech. This assemblage, although poor in numbers and diversity, contains
elements suggestive of north-temperate climates, i.e. N. labyrinthus and P. pentagonum , and is
possibly interpretable as assignable to the Allerod/Bolling Stade (see later). Between 0-4 and 01 m
depth there is a second group of samples that yield dinoflagellate cyst assemblages. These
assemblages give dinoflagellate cyst recovery figures of up to 1100 cysts per gram of sediment and
are dominated by O. centrocarpum , accompanied by N. labyrinthus , P. pentagonum , Spiniferites
elongatus Reid, and S. mirabilis (Rossignol) Sarjeant together with minor amounts of round, brown
Protoperidinium cysts. The curve of the cyst numbers clearly shows a rising trend closely
corresponding to the upward shift to progressively oxygen lighter water as the global ice volume
decreased. As with the sample at 0-9 m, these assemblages are also interpreted as being north-
temperate in nature and indicating environments similar to today. All three of these uppermost
samples can, therefore, be assigned to the Holocene (see later).

268

PALAEONTOLOGY, VOLUME 37

DSDP HOLE 552 A

text-fig. 2. Dinoflagellate cyst spectrum for DSDP Hole 552A. X axis in thousands of cysts per gram. P. spp
(RB) indet. indicates the counts of indeterminate round, brown Protoperidinium cysts.

table 2. DSDP Hole 552A — dinoflagellate cysts per gram.

Depth (m) ...

010

018

0-30

0-39

0 51

0-62

0-71

0-89

100

Gonyaulacaceae

B. tepikiense

—

—

—

—

—

—

—

56

—

N. labyrinthus

78

47

—

—

—

—

—

1 1

—

0. centrocarpum

721

442

—

—

—

—

—

67

—

S. elongatus

78

79

—

—

—

—

—

—

—

S. mirabilis

117

63

34

—

—

—

—

—

—

Peridiniaceae

P. conicoides

—

16

—

—

—

—

—

—

—

P. pentagonum

38

47

17

—

—

—

—

11

—

P. spp. (RB)

—

47

—

—

—

—

—

—

—

n

1033

741

51

—

—

—

—

145

—

DSDP Hole 610A

This hole was drilled as part of DSDP Leg 94 on the western side of the Rockall Trough at the crest
of the Feni Ridge. It was drilled using the HPC and provided detail of the Neogene and Quaternary
history of the ridge. Recently acquired and unpublished stable isotope data from Professor Eystein
Jansen are included in Text-figure 3 with his permission. Unfortunately, there is obvious
transportation and disturbance of the sediment at the top of the Hole giving glacial values (Jensen
in lift.). Consequently, it appears that although Termination IA might occur at the base of the
sequence studied, both Termination IB and the subsequent record are too disturbed to be
recognized. Also, there are no detailed accounts of the sedimentology of the core so it is not possible
to state if volcanic ashes are present. It has thus proved difficult to place the sequence in a time
framework. However, the total dinoflagellate cyst recovery is good and appears not to have been
affected by any disturbance; the reason for this is unknown. If the same chronostratigraphical
assumptions are made, a maximum sedimentation rate of 150 mm per Ka for the Younger Dryas
and about 80 mm per Ka for the Holocene follows. However, as with the record of 552A, it is likely
that much of the later Holocene is missing.

Given that these assumptions are correct, the four main chronostratigraphical divisions, as
illustrated in Text-figure 3, are characterized by different dinoflagellate cyst assemblages. All the
dinoflagellate cyst data are given in Table 3. The assemblages are as follows; the latest glacial yields
poor numbers of individuals but contains some B. tepikiense and O. centrocarpum; the

HARLAND: NORTH ATLANTIC DINOFLAGELLATES

269

text-fig. 3. Dinoflagellate cyst spectrum for DSDP Hole 610A. Legend as for Text-figure 2.

table 3. DSDP Hole 610A — dinoflagellate cysts per gram.

Depth (m) ...

0-12

0-30

0-40

0-60

0-75

0-90

106

1 20

1-35

1-48

Gonyaulacaceae

B. tepikiense

17

—

—

—

—

111

471

78

12

24

l. aculeatum

—

—

—

—

—

—

13

—

—

—

I. paradoxum

—

21

—

—

—

—

—

—

—

—

N. labyrinthus

182

717

748

1385

Ill

100

—

—

—

—

O. centrocarpum

1358

1209

888

399

78

122

178

22

38

—

S. e/ongatus

133

62

187

141

—

22

—

—

—

S. mirabilis

50

164

257

47

—

—

—

—

—

—

S. ramosus

—

—

—

—

—

1 1

—

—

—

S. spp indet.

33

62

—

—

—

—

13

—

—

—

Peridmiaceae

P. pentagonum

33

21

—

—

—

—

—

—

—

—

P. spp (RB)

—

—

—

—

—

—

38

—

—

—

n

1806

2256

2080

1972

189

366

715

100

50

24

Allerod/ Bolling has sharply increased numbers of cysts to c. 650 cysts per gram, consisting mainly
of B. tepikiense with small numbers of O. centrocarpum , round, brown Protoperidinium cysts and
Impagidinium species, together with the first appearance of N. labyrinthus and S', e/ongatus; the
Younger Dryas has low numbers of cysts but includes B. tepikiense , N. labyrinthus , O. centrocarpum
and S. e/ongatus ; and finally the Holocene with cyst numbers in excess of 2200 cysts per gram and
characterized by high numbers of N. labyrinthus and O. centrocarpum with Impagidinium spp.,
P. pentagonum, and Spiniferites cysts. Nematosphaeropsis labyrinthus exhibits a distinct peak
abundance at 0-6 m at about the interpreted level of Termination IB.

Unfortunately the chronostratigraphy for DSDP Hole 610A has proved somewhat elusive,
although the dinoflagellate cyst record appears to be particularly clear and more complete than that
of Hole 552A.

BGS Gravity Core 56/ -10/ 47

This gravity core was taken as part of the BGS survey of the British continental shelf and in
particular to the production of the 1 : 250000 Quaternary Geology edition of the Peach Sheet (James
1991). Unfortunately no stable isotope or detailed sedimentology has been attempted on this core
to date, despite its obvious interest. It is anticipated that some tephrachronological results may be
available in future. However, for the moment the detailed dinoflagellate cyst analyses must stand

270

PALAEONTOLOGY, VOLUME 37

alone. The chronostratigraphy outlined in Text-figure 4 is. therefore, wholly interpretational based
partly upon comparisons with the two DSDP Holes, discussed earlier, where there is some limited
chronostratigraphy. A maximum interpreted sedimentation rate of 260 mm per Ka may have
operated during the Younger Dryas, but the rate was more likely to have been about 50 mm per
Ka for the Holocene. It is particularly clear that the record obtained from the dinoflagellate cysts
is complete and well demonstrated within this gravity core. The dinoflagellate cyst spectrum is
illustrated in Text-figure 4 and the data are given in Table 4.

BGS GRAVITY CORE 56/-10/47

text-fig. 4. Dinoflagellate cyst spectrum for BGS Gravity Core 56/-10/47. Legend as for Text-fig. 2. P. spp.
(P) indet. and P. spp. (RB) indet. refer to the indeterminate peridinioid and round, brown Protoperidinium
species respectively. 1, Pre-Boreal; 2. Boreal; 3, Atlantic; 4, Sub-Boreal; and 5, Sub-Atlantic.

The Late Glacial part of the sequence is characterized by poor recovery of cysts ( < 200 cysts per
gram of sediment) but contains B. tepikiense and round, brown Protoperidinium cysts. The
Allerod/Bolling Interstade, in contrast, demonstrates a marked increase in cyst recovery (> 500 to
> 2000 cysts/gram) with high numbers of B. tepikiense and lesser numbers of round, brown
Protoperidinium cysts, together with Impagidinium spp., P. conicum and P. pentagonum.
Operculodinium centrocarpum is also notably present, coming to a peak of recovery slightly later
than that of B. tepikiense ; this undoubtedly has some oceanographic significance (see later). The
Younger Dryas demonstrates a return to low recovery (< 200 cysts per gram) but is characterized
by roughly equal amounts of O. centrocarpum and B. tepikiense together with round, brown
Protoperidinium cysts. Finally the Holocene part of the interpreted sequence is characterized by
markedly high cyst recovery (> 1000 to 10000 cysts per gram) with particularly high numbers of
O. centrocarpum and N. labyrinthus. In addition Spiniferites species such as S. elongatus and
5. mirabilis also show marked increases as do Protoperidinium conicum , P. pentagonum and round,
brown Protoperidinium cysts.

This sequence is probably the most complete and full dinoflagellate cyst record for this part of
the northeastern Atlantic and potentially offers a key to the interpretation of the palaeoceanography
through the last deglaciation. A full discussion is given later.

BGS Vibrocore 57/-10/84

The second BGS core was also taken as part of the BGS survey of the United Kingdom continental
shelf and, in particular, to the production of the 1 : 250000 Quaternary Geology St Kilda Sheet
(Evans 1992). Although no stable isotope work is available, the core was analysed in detail for ash
shards (Selby 1989) and, over the interval under discussion, two ash layers were discovered. The first
downhole was at 0 60 m and the second at 0 90 m; these ashes may be correlated with the Vedde
Ash at 10600 YBP and an earlier fall not previously recognized (Selby 1989). Assuming that the
chronostratigraphy is approximately correct, the sequence is interpreted as in Text-figure 5. This
implies a maximum sedimentation rate of about 1 m/Ka during the Younger Dryas and about
60 mm per Ka for the Holocene. It seems likely that parts of the Holocene are missing (see later).

The dinoflagellate cyst record for 57/- 10/84 is illustrated in Text-figure 5 and documented in

table 4. BGS Gravity Core 56/-10/47 — dinoflagellate cysts per gram.

HARLAND: NORTH ATLANTIC DINOFLAGELLATES

271

I I

I I I

I I rl

I C4

I I

I £ I

I of (N M
I 04 I — co

I I

CO | c<r VO

I - I -

I I Pi

I I I I

I I

INI!

I I I

I I I

I I I

I !

r- | os 04
04 | vo >o oo

oo Os

— — 04

I I

o o o r- o | 04 —
- ^ inn - | oo o
r- — — m

•— < co

i co VO C l h iti CO O

oo vo oo m oo n oo

VO 04 of 04 04

CO 04

3 I

ct ti iri o

04 — < of
of 0 4 OO —
0 4 VO)

I WO VO

I of —

I

I 3!2!

I 8 I

I I

os I Os of — ■
04 I t-~ co os
m o 04

of — - I G 1

O Os I — i

0 4 Os — 1

— 5,
3 Sr

VJ Q Q

&1

£ t>0 ^ ~ ^

c c 3 s: ?

o Li ^

^^OCoCoCoCoCo

_ co
cu oC

• g ^ CL,
<u

CO

O O C 3
C Ci Cl,

■ C^

Cl. Cl,

272

PALAEONTOLOGY, VOLUME 37

BGS VIBROCORE 57/-1 0/84

text-fig. 5. Dinoflagellate cyst spectrum for BGS Vibrocore 57 /— 1 0/84. Legend as for Text-figure 2.

detail in Table 5. Pre Allerod/Bolling sediments appear to have not been penetrated but the
interstade is characterized by high cyst recovery ( > 500 to 1600 cysts per gram of sediment) and the
dominance of B. tepikiense , although O. centrocarpum , N. labyrinthus , and round, brown
Protoperidinium cysts together with Impagidinium spp. are also present. In contrast, the Younger
Dryas has low cyst recovery (< 200 cysts/gram) but with B. tepikiense and round, brown
Protoperidinium species and some O. centrocarpum. The Holocene contains much higher numbers
of cysts (often > 3000 cysts per gram) with high numbers of N. labyrinthus and O. centrocarpum ,
together with an influx of Impagidinium spp., Protoperidinium conicum , P. pentagonum and
Spiniferites spp. including S. elongatus and S. mirabilis.

SYNTHESIS

The dinoflagellate cyst record across the glacial/postglacial transition is similar in all four cores
studied and a tentative synthesis can be made. It is convenient to attempt this in a series of time
slices mimicking those used for the deglaciation history (Ruddiman 1987). Fortunately there is
sufficient consistency between the studied sites to make this approach realistic; marked differences
thought to be of oceanographic significance are noted. The dinoflagellate cyst assemblages, their
interpretations and the climatic and oceanographic implications are given below.

Late Glacial > 13000 YBP

Off the western coast of Scotland on the continental slope, the dinoflagellate cyst record from BGS
Gravity Core 56/-10/47 for this time slice consists of a low diversity, poor recovery assemblage (see
Table 4). Species present mostly include B. tepikiense , round, brown Protoperidinium cysts, together
with minor amounts of O. centrocarpum. Certainly B. tepikiense contributed most to this
assemblage. In BGS Vibrocore 56/-10/84 (time slice not illustrated here) the situation is similar to
that noted above (Harland, unpublished data ) with, in addition, occasional specimens of
? Algidasphaeridium minutum (Harland and Reid) Matsuoka and Bujak. Similar assemblages have
also been described by Peacock et al. (1992) for Late Glacial sediments recovered in vibrocores,
dated at > 15245 YBP, on the shelf area west of Scotland and also in similarly dated sediments,
> 12785 YBP, from a nearshore situation in the North Minch (Graham et al. 1990).

Further out into the north-eastern Atlantic Ocean, in the proximity of the Rockall Plateau,
sediments of this age proved barren of dinoflagellate cysts from DSDP Hole 552A but DSDP Hole
610A yielded similar assemblages to those described above. Assemblages for the latter site are.

HARLAND: NORTH ATLANTIC D1NOFLAGELLATES

273

274

PALAEONTOLOGY, VOLUME 37

therefore, like those on the continental slope and shelf. Furthermore, the dinoflagellate cyst
assemblage described from Unit 1 of Stoker et al. (1989) from a vibrocore in the Faeroe-Shetland
Channel is also comparable but contains rather more round, brown Protoperidinium cysts.
However, it is difficult, if not impossible, to assign all these assemblages unequivocally to exactly
the same time interval and so, therefore, they may provide information only on particular
environments within the Late Glacial.

However, undoubtedly within this time the marine sediments provided a characteristic low
diversity and low recovery flora, predominantly containing B. tepikiense and round, brown
Protoperidinium cysts. This type of cyst assemblage is known to characterize cold, arctic-like
environments with a minimum of North Atlantic Current influence (see Wall et al. 1977; Harland
1983; Mudie and Short 1985). The presence of round, brown Protoperidinium cysts may indicate the
possibility of sea-ice (Dale 1985) and high numbers of B. tepikiense possible meltwater influxes
lowering the sea-water salinity. These interpretations are based on increasing knowledge of the
ecology of modern dinoflagellate cysts. Additionally these sediments often contain high proportions
of reworked palynomorphs indicative of the high levels of erosion from shelf areas (see Stoker
et al. 1989).

Throughout this time, the area under consideration was entirely glacial, with little NAC influence
and probably often near sea-ice. This compares well with the scenario illustrated by Ruddiman
(1987).

AUerod/ Bolling Interstade 13000-1 1 000 YBP

This time slice includes the beginning of deglaciation in the North Atlantic Ocean. In the two BGS
cores off the west coast of Scotland, this is indicated by both a sharp rise in diversity and in cyst
recovery. In 56/-10/47 the rise in cyst recovery is somewhat stepped before reaching a peak of over
2000 cysts per gram of sediment. Text-figure 4 shows the clear predominance and importance of
B. tepikiense and, to a lesser extent, the later influx of O. centrocarpum. The additional presence
of Impagidinium spp., N. labyrinthus , Protoperidinium conicum and P. pentagonum is also of
importance. This situation is almost exactly mirrored in 56/- 10/84 but without the O. centrocarpum
peak and the presence of the two Protoperidinium species. Although not recognized in shelf
sediments, the Allerod/Bolling Interstade was also noted in the nearshore sediments of the North
Minch (Graham et al. 1990); here the same interval is represented by units D2 and D3, the former
dominated by Protoperidinium cysts with considerable amounts of B. tepikiense , and the latter by
high proportions of O. centrocarpum. Also consistently present are Spiniferites cysts such as S. lazus
Reid and 5. ramosus (Ehrenberg) Loeblich and Loeblich. This shows a remarkable compatability
with the offshore record but contains more elements consistent with a nearshore shelf situation.
Units D2 and D3 are conveniently constrained by 14C dates and are confidently assignable to the
Windermere Interstade.

In the vicinity of the Rockall Plateau this time slice is characterized by a single sample in DSDP
Hole 552A, and a rather better record in DSDP Hole 610A consisting of an increase in cyst diversity
and recovery to a peak of 715 cysts per gram of sediment. The assemblage is dominated by
B. tepikiense but also contains round, brown Protoperidinium cysts, P. pentagonum and some
Spiniferites cysts together with N. labyrinthus towards the top. The basic dinoflagellate cyst
signature, however, appears to be similar to. that on the continental slope and in the nearshore area.
Further north in the Northern Rockall Trough and the Faeroe-Shetland Channel, the dinoflagellate
cyst assemblages of Unit 2 (Stoker et al. 1989) are somewhat different. Here the cyst flora remains
dominated by Protoperidinium species, particularly round, brown cysts, with significant
O. centrocarpum and other minor cysts. Although some amelioration is evident, rather cold and
severe conditions nevertheless prevailed, possibly with the proximity of sea-ice. This may indicate
the extent of retreat of the polar front during this time to a position on the Wyville-Thomson Ridge.

The Allerod/Bolling time slice, like that of the Late Glacial, yields a reasonably homogenous
dinoflagellate cyst assemblage. This assemblage is dominated by B. tepikiense which may be

HARLAND: NORTH ATLANTIC DINOFLAGELLATES

275

interpreted as indicating the release of large quantities of freshwater into the marine environment
as the ice retreated. This caused a general lowering of the salinity of the surface water in the north-
eastern Atlantic Ocean. Since this phenomenon is recorded in the deep ocean, on the continental
slope and on the shelf, it must have been a widespread and significant oceanographic event. Further
north, in the vicinity of the Wyville-Thomson Ridge, the dinoflagellate cyst assemblages are
indicative of more severe climatic conditions with similarities to the earlier Late Glacial, but
showing some influence from the North Atlantic. Towards the later stages of the Allerod/Bolling,
increased NAC influence is evidenced by the increased amounts of O. centrocarpum and
N. labyrinthus. The implication is that the polar front, or at least sea-ice, remained in the north-
eastern Atlantic until towards the end of the interstade. It is doubtful if any of the records discussed
herein are complete, so some caution must be exercised in the interpretation of these results.
Nonetheless it does seem clear that sea-ice remained in the area for much of the time, contributing
to the freshwater input into the system and influencing the cyst assemblages. This may not accord
in detail with Ruddiman (1987), but it may explain part of the feedback mechanism that released
large quantities of freshwater, adding to that entering from the Laurentide ice sheet (Broecker et al.
1988, 1989), and returning the Atlantic to a more glacial scenario. The evidence of Baumann and
Matthiessen (1992), from the Norwegian Sea at this time, also suggests that any influx of North
Atlantic water would have been diluted by large volumes of meltwater.

Younger Dry as 11 000-10000 YBP

The Younger Dryas comprises an enigmatic return to cold climates following the initiation of the
deglacial cycle and is the subject of much controversy.

The two continental slope cores off western Scotland revealed a marked and sudden decline in
both cyst diversity and recovery. The assemblages return to those dominated by round, brown
Protoperidinium cysts and B. tepikiense , together with some O. centrocarpum and N. labyrinthus.
BGS Vibrocore 56/- 10/84 appears to prove a thicker sequence of Younger Dryas sediments but the
overall assemblage characteristics are the same as that noted above with, perhaps, further detail of
the temporal changes in the cyst flux to the sediment. Of interest is the initial peak of B. tepikiense
and a final peak of O. centrocarpum but without collaborative evidence it would be unwise to
speculate further. In the nearshore North Minch Borehole 78/4 (Graham et al. 1990) Unit D4 is
interpreted as of Younger Dryas age as indicated by radiocarbon dating. The dinoflagellate cyst
assemblages are dominated by round, brown Protoperidinium cysts but also contain O. centrocarpum
and Spiniferites spp. such as S. lazus and S. ramosus.

Further offshore in the vicinity of the Rockall Plateau, DSDP Hole 552A is barren of
dinoflagellate cysts and DSDP Hole 610A appears to have an attentuated sequence. However, the
assemblages are low in diversity and recovery containing O. centrocarpum and N. labyrinthus. To
the north. Unit 3 of Stoker et al. (1989) in both the North Rockall Trough and the Faeroe-Shetland
Channel are characterized by B. tepikiense with O. centrocarpum and Spiniferites cysts.

The interpretation of the Younger Dryas dinoflagellate cyst record is difficult. Without the
knowledge and confidence that a full and complete sequence is available, much of the interpretation
must be speculative. The cyst assemblages from the continental slope and the nearshore area
certainly suggest a return to glacial-like conditions but there are sufficient numbers of temperate
cysts of North Atlantic affinity to indicate a difference between this cold interval and that of the Late
Glacial at > 13000 YBP. Perhaps the Atlantic Ocean was functioning more like the present than
during full glacial times as Jensen and Veum (1990) have suggested. Perhaps the Atlantic Ocean
oceanography did not fully return to its previous state and that some vestige of a NAC remained,
feeding some warmer water to the higher latitudes.

276

PALAEONTOLOGY, VOLUME 37

Holocene 10000 YBP - Present

A major change in the dinoflagellate cyst assemblages, even greater than that at the initiation of the
Allerod/ Bolling Interstade, heralds the Holocene. In previous publications this transition has been
described in a number of different ways including 'event 3’ of Harland (1984) and the change from
B. tepikiense-dominated assemblages to those characterized by O. centrocarpum (Harland 1988). In
both instances, the detail of the transition from the Late Glacial to the Post Glacial was not
available. However, Turon (1980, 1981) had earlier documented in detail this change in the
dinoflagellate cyst assemblages across the Pleistocene/Holocene boundary in cores from the
Rockall Channel.

In the present study, the two DSDP cores clearly show typical Holocene assemblages
characterized by high diversity and high cyst recovery to, in some cases, over 2000 cysts per gram
of sediment. The assemblages, as intimated above, are usually characterized by O. centrocarpum
together with N. labyrinthus and often contain such Spiniferites cysts as 5. elongatus and S. mirabilis
together with Protoperidinium conicum and P. pentagonunv, cyst species such as B. tepikiense and
round, brown Protoperidinium cysts are rarer. DSDP Hole 610A contains a particularly good
Holocene dinoflagellate cyst spectrum that portrays an initial sharp rise in N. labyrinthus before its
decline and a subsequent rise in O. centrocarpum. In other words, there is a definite pattern within
the Holocene of the numbers of cyst species being incorporated into the bottom sediments. This is
also observed within the less conspicuous and less numerous members of the assemblage such that,
for instance, the Spiniferites species rise in numbers toward the later stages of the N. labyrinthus
peak.

Closer inshore, within the two BGS cores taken on the continental slope, the Holocene appears
to be much more complete and furnishes a more detailed dinoflagellate cyst spectrum. This is
especially true for 56/-10/47 (Text-figure 4) where cyst recovery reaches over 10000 cysts per gram
of sediment. The Holocene spectrum can be divided into a number of phases largely based upon the
relative abundances of O. centrocarpum and N. labyrinthus. These phases are listed below.

Phase 1

Cyst numbers began to increase sharply from the low recovery of the Younger Dryas and contain
O. centrocarpum together with round, brown Protoperidinium cysts, in addition to increasing
numbers of S. elongatus , .S', mirabilis, P. conicum, P. pentagonum and Impagidinium cysts. This phase
may be present within the record of DSDP Hole 552A but does not appear to be present in Hole
610A.

Phase 2

This is represented by a distinctive peak in the N . labyrinthus curve as the general cyst recovery
improves. It may be present within 610A but is certainly absent in 552A; the recognition of these
phases is difficult and often dependent upon sampling interval and sedimentation rate.

Phase 3

This coincides with the maximum cyst recovery in 56/-10/47 and peaks in the curves of
O. centrocarpum, S. elongatus, S. mirabilis, round, brown Protoperidinium cysts and P. conicum. It
may also be present in 610A but not in 552A.

Phase 4

This is recognized by the return of a peak in the N. labyrinthus curve and some decrease in recovery
of other cyst species. However, this reduction in cyst numbers is relatively small and certainly not
of the order of those seen in the older glacial sediments. This phase does not seem to be present in
either of the two DSDP holes.

UARLAND: NORTH ATLANTIC D I N OFL AG E L L AT ES

277

Phase 5

This final phase comprises a rise in the cyst recovery and also in the curves of O. centrocarpum,
S. elongatus , round, brown Protoperidinium cysts, P. conicum and P. pentagonum. This is
accompanied by a decline in N. labyrinthus and S. mirabilis. Like the phase described above, this
also does not appear to be present in the two DSDP holes.

This description relies heavily upon 56/-10/47 as a standard for the area and assumes that it has
sampled a complete sequence, or at least the most complete sequence of the Holocene known to date
from the offshore area; for the moment it has not been possible to test this assumption.

In BGS core 57/-10/84 it is possible to recognize a number of the phases described above. In
particular, it would appear that phases 4 and 5 are present but that some of the older phases are
not. However, there are certain differences between the two that complicate the issue, including the
loss of cyst recovery at a mid-point in the sequence, and the coincidence of an O. centrocarpum and
a B. tepikiense peak in the lower part of the sequence.

All the Holocene dinoflagellate cyst assemblages outlined above are similar, with the exception
of increased numbers of shelf species in the BGS cores, as might be expected from their location.
Otherwise, it would appear that the climatic and oceanographically controlled assemblages occur
throughout the area and offer the potential of detailed correlation. The sequences of the Northern
Rockall Trough and the Faeroe-Shetland Channel (Stoker et al. 1989) encompassed within Unit 4
are typically characterized by rich cyst recovery and dominated by O. centrocarpum. It is, however,
difficult to place the assemblages with respect to the phases reported herein, as the sequence is only
some 0-45 m thick and the data were collected as percentages and not as absolute numbers.

In contrast, the sequence from the North Minch (Graham et al. 1990) proved over 16 m of
Holocene that was divisible into three units and four subunits. It is, therefore, attractive to regard
these as directly correlatable with the five phases described above. However, the North Minch
Holocene assemblages, confirmed by radiocarbon dating, are dominated by Spiniferites species
especially S. lazus , S. mirabilis and S. ramosus , with O. centrocarpum being a minor component only
and N. labyrinthus being absent. Only the use of chronostratigraphical methods can establish the
equivalence of these two schemes but undoubtedly the potential of using dinoflagellate cyst analysis
for these high-resolution investigations is demonstrated herein.

COMPARISONS AND DISCUSSION

The pioneering work of Turon (1980, 1981) on the Rockall Channel revealed similar dinoflagellate
cyst spectra to those described herein. In particular, they displayed the same prominence and
sequential arrangement of peak occurrences of the species B. tepikiense , N. labyrinthus and
O. centrocarpum (see especially the record for core 73136). Differences in Turon’s spectra are mainly
confined to the occurrence of high percentages of Impagidinium species, particularly I. sphaericum
in the Holocene parts of the sequences. This difference may reflect the offshore nature of Turon’s
sites and the increased oceanic nature of the environment of deposition; Impagidinium spp. are well
known indicators of the oceanic realm (Wall et al. 1977). In addition, Turon’s cores contained
increased percentages of B. tepikiense in the late Holocene sediments; a situation not recognized
here but may be explained by the transportation of cysts in cold bottom water currents, as the
present oceanographic configuration became firmly established. Turon (1978) proposed that the
changes in the Holocene dinoflagellate cyst assemblages might reflect differences in primary
productivity. Such differences were linked to the availability of nutrients caused by alterations in
storm tracks across the North Atlantic Ocean affecting the oceanography.

More recently, De Vernal et al. (1992) reviewed the dinoflagellate cyst record for Quaternary
sediments from the North Atlantic and, in their discussion of short-term high-resolution data,
included those of Turon (1980, 1981) and that of BGS Gravity Core 56/-10/47 discussed in detail
herein. These authors pointed out that changes in the dinoflagellate cyst assemblages could effect
direct and accurate ecostratigraphical correlations across the region that reflected synchronous, or

278

PALAEONTOLOGY, VOLUME 37

almost synchronous, changes in the environment. These changes were oceanographic in nature and
occurred as a result of climate-forcing. In particular, marked oceanographic alterations are known
to have occurred as the polar front moved across the area during deglaciation.

Mattiessen (1991 ), in work published on the dinollagellate cysts of the Norwegian Sea, suggested
that the North Atlantic has influenced the area since about 15 Ka. Modern circulation patterns were
initiated around 10 Ka, with N. labyrinthus dominating the cyst assemblages, until between 6 and
7 Ka when the present oceanography was fully established. Further work by Baumann and
Mattiessen ( 1992), utilizing both dinoflagellate cysts and coccoliths, established several distinct steps
in Holocene oceanography, not unlike those discussed earlier for the north-eastern Atlantic. In
particular, after the first initiation of the surface water circulation, slightly cooler water conditions
are thought to have prevailed followed by a major change at the time of the climatic optimum
(c. 6000 YBP) as the present hydrography became established; they also suggested that there is
some evidence for a decrease in sea-surface temperatures since about 4000 YBP.

All previous work, and that described here, are similar in respect to the deglaciation history of
the north-eastern Atlantic from the evidence furnished by dinoflagellate cyst assemblages. The
initial phase, marked by Termination IA in the oxygen isotope signature, is uniquely characterized
by the occurrence of high percentages of B. tepikiense , a cyst species known to favour north
temperate to arctic environments and less than fully marine salinities. Although the NAC was
probably active at this stage, it is thought likely that large quantities of meltwater were entering the
system and effectively lowering the sea-surface salinity. This phenomenon itself might well have
been sufficient to reduce the flow of NADW and hence trigger the return of conditions akin to the
full glacial situation. The cyst assemblages recovered from Younger Dryas sediments, although low
in numbers and diversity, do contain species associated with the activity of the NAC; they are not
the same as those recovered from sediments associated with full glacial environments.

The dinoflagellate cyst assemblages obtained from the Holocene sediments have proved not to be
uniform, but to show distinct changes in character. These changes echo those already documented
by Turon (1980, 1981), De Vernal et a/. (1992) and Baumann and Mattiessen (1992). Even after
deglaciation had been achieved, major changes were occurring in the oceanography of the North
Atlantic Ocean and Norwegian Sea. Despite proposals published by Turon (1978) and Baumann
and Mattiessen (1992) as to the likely changes and their causes, it is unfortunately true that there
are insufficient ecological data available to interpret the assemblages with any degree of confidence.
However, palynologists working with pollen diagrams have long been able to subdivide their
Holocene spectra based upon changes in climate. It is likely that both the fluctuations in pollen
diagrams and in the dinoflagellate cyst assemblages are related, and may have the potential to assist
in the elucidation of Holocene oceanographic and climate change. Perhaps notions of fluctuations
in the position of the atmospheric jet-stream, the pathways of anticyclones and the alternation of
dry and wet climates, first recognized in Scandinavia (Sernander 1908), together with changes in the
oceanography of the North Atlantic and Norwegian Sea are all inextricably linked.

This paper has demonstrated that the study of dinoflagellate cyst assemblages through the last
deglaciation gives an insight into changes that are occurring in the surface waters of the ocean at
a time when the globe is moving from a glacial climate to that of the present day. However, there
is a distinct need for additional ecological information on both dinoflagellates cysts and the
individual species, to aid the interpretation of the recovered assemblages. In this respect the work
of Dale and Dale (1992) is worth consideration, as it is the only available study that attempts to
examine the nature of the dinoflagellate cyst flux to the bottom sediments.

Acknowledgements. This study is a result of a BGS Research and Development Programme entitled
‘ Dinoflagellate cyst climatostratigraphic synthesis of the Quaternary of the north-eastern Atlantic Ocean ’. The
work would have been impossible without the assistance of Ms Jane Kyffin-Hughes whose careful preparations
have served as the foundation for the qualitative and semi-quantitative results. Especial thanks are due to
Professor Eystein Jansen for providing both the samples and oxygen isotope analyses for DSDP Hole 610A,

HARLAND: NORTH ATLANTIC DINOFLAGELLATES

279

and for showing an interest in the project. Dr J. B. Riding kindly read and offered constructive criticisms on
an earlier draft of this paper. Publication is with permission from The Director, British Geological Survey
(NERC).

REFERENCES

balech, E. 1974. El genero ‘Protoperidinium’ Bergh, 1881 (‘Peridinium’ Ehrenberg, 1831, partim). Revista del
Museo Argentino de Ciencias Naturales ‘ Bernardino Rivadavia Hidrobiologta , 4, 1-79.
bard, e., Arnold, m., Maurice, p., duprat, j., moyes, j. and duplessy, j.-c. 1987. Retreat velocity of the North
Atlantic polar front during the last deglaciation determined by UC accelerator mass spectrometry. Nature ,
328. 791-794.

baumann, k.-h. and matthiessen, J. 1992. Variations in surface water mass conditions in the Norwegian Sea:
evidence from Holocene coccolith and dinoflagellate cyst assemblages. Marine Micropaleontology , 20,
129-146.

berger, w. H. 1990. The Younger Dryas cold spell - a quest for causes. Palaeogeography , Palaeoclimatology ,
Palaeoecology , 89, 219-237.

boyle, e. a. and keigwin, l. 1987. North Atlantic thermohaline circulation during the past 20,000 years linked
to high-latitude surface temperature. Nature , 330, 35-40.
broecker, w. s., andree, m., wolfli, w., oeschger, h., bonani, g., kennett, j. and peteet, d. 1988. The
chronology of the last deglaciation : implications to the cause of the Younger Dryas event. Paleoceanography ,
3, 1-19.

— kennett, j. p., flower, b. p., teller, j. t., trumbore, s., bonani, g. and wolfli, w. 1989. Routing of
meltwater from the Laurentide Ice Sheet during the Younger Dryas cold episode. Nature , 341, 318-321.

butschli, o. 1873. Einiges iiber Infusorien. Archiv fur mikroskopische Anatomie , 9, 657-678.
cline, r. m. and hays, J. d. (eds). 1976. Investigation of Late Quaternary paleoceanography and
paleoclimatology. Memoirs of the Geological Society of America , 145, 1^164.
dale, a. l. and dale, b. 1992. Dinoflagellate contributions to the sediment flux of the Nordic Seas. Woods Hole
Oceanographic Institution , Ocean Biocoenosis Series , 5, 45-76.
dale, b. 1983. Dinoflagellate resting cysts: ‘benthic plankton’. 69-136. In fryxell g. r. (ed. ). Survival
strategies of the algae. Cambridge University Press, Cambridge, 154 pp.

— 1985. Dinoflagellate cyst analysis of Upper Quaternary sediments in core GIK 15530-4 from the
Skagerrak. Norsk Geologisk Tidsskrift , 65, 97-102.

dansgaard, w., white, j. w. c. and johnsen, s. j. 1989. The abrupt termination of the Younger Dryas climate
event. Nature , 339, 532-534.

denton, g. h. and karlen, w. 1973. Holocene climatic variations; their pattern and possible cause. Quaternary
Research , 3. 155-205.

duplessy, j.-c., Arnold, m., Maurice, p., bard, e., duprat, j. and moyes, j. 1986. Direct dating of the oxygen-
isotope record of the last deglaciation by 14C accelerator mass spectrometry. Nature , 320, 350-352.
edwards, l. e., mudie, p. j. and vernal, a. de 1991. Pliocene paleoclimatic reconstruction using dinoflagellate
cysts: comparison of methods. Quaternary Science Reviews , 10, 259-274.
ehrenberg, c. g. 1832. Beitrage zur K.enntnis der Organisation der Infusorien und ilirer geographischen
Verbreitung, besonders in Sibirien. Abhandlungen der Preussischen Akademie der Wissenschaften, 1830, 1-88.
Evans, c. D. R. 1992. St. Kilda (57 °N-I0 °W) Quaternary Geology. 1:250000 Offshore Map Series. British
Geological Survey.

evitt, w. r. 1985. Sporopoltenin dinoflagellate cysts. Their morphology and interpretation. American Association
of Stratigraphic Palynologists Foundation, Austin, 333 pp.

Fairbanks, R. G. 1989. A 17,000-year glacio-eustatic sea level record: influence of glacial melting rates on the
Younger Dryas event and deep-ocean circulation. Nature, 342, 637-642.
fritsch, f. e. 1929. Evolutionary sequence and affinities among Protophyta. Biological Review , 4, 103-151.
graham, d. k., harland, r., Gregory, d. m., long, d. and morton, a. c. 1990. The biostratigraphy and
chronostratigraphy of BGS Borehole 78/4, North Minch. Scottish Journal of Geology, 26, 65-75.
haeckel, E. 1894. Entwurf eines natiirlichen Systems der Organismen auf Grand Hirer Stammegeschichte . Georg
Reimer, Berlin, 400 pp.

harland, r. 1982. A review of Recent and Quaternary organic-walled dinoflagellate cysts of the genus
Protoperidinium. Palaeontology , 25, 369-397.

— 1983. Distribution maps of Recent dinoflagellate cysts in bottom sediments from the North Atlantic
Ocean and adjacent seas. Palaeontology , 26, 321-387.

280

PALAEONTOLOGY, VOLUME 37

harland, r. 1984. Quaternary dinoflagellate cysts from Hole 552A, Rockall Plateau, Deep Sea Drilling Project
Leg 81. Initial Reports of the Deep Sea Drilling Project , 81, 541-546.

1988. Dinoflagellates, their cysts and Quaternary stratigraphy. New Phytologist , 108, 111- 120.

— 1989. A dinoflagellate cyst record for the last 0-7 Ma for the Rockall Plateau, northeast Atlantic Ocean.
Journal of the Geological Society , London , 146, 945-951.

- 1992a. Dinoflagellate cysts of the Quaternary System. 253-273. In powell, a. j. (ed. ). A stratigraphic
index of dinoflagellate cysts. Chapman and Hall, London, 290 pp.

- 19926. Dinoflagellate cyst biostratigraphy of the last 2-3 Ma from the Rockall Plateau, northeast Atlantic
Ocean. Journal of the Geological Society, London, 149, 7-12.

James, J. w. c. 1991. Peach ( 56°N-10°W ) Quaternary Geology. 1:250000 Offshore Map Series. British
Geological Survey.

jansen, e. and veum. t. 1990. Evidence for two-step deglaciation and its impact on North Atlantic deep-water
circulation. Nature, 343, 612-616.

keigwin, l. d. and jones, G. a. 1989. Glacial-Holocene stratigraphy, chronology, and paleoceanographic
observations on some North Atlantic sediment drifts. Deep-Sea Research, 36, 845-867.
rofoid, c. a. and swezy, o. 1921. The free-living unarmored Dinoflagellata. Memoirs of the University of
California, 5, 1-562.

kudrass, h. r., erlenkeuser, h., vollbrecht, r. and weiss, w. 1991. Global nature of the Younger Dryas
cooling event inferred from oxygen isotope data from Sulu Sea cores. Nature, 349, 406^109.
kvamme, t., mangerud, j., furnes, h. and ruddiman, w. f. 1989. Geochemistry of Pleistocene ash zones in
cores from the North Atlantic. Norsk Geologislc Tidsskrift, 69, 251-272.

Lehman, s. j. and keigwin, l. d. 1992. Sudden changes in North Atlantic circulation during the last
deglaciation. Nature, 356, 757-762.

lemmermann, e. 1910. Kryptogamenfiora der Mark Brandenburg and angrenzender Gebiete, III Algen I
(Schizophyceen, Flagellaten, Peridineen ), Gebriider Borntraeger, Leipzig, 120 pp.
lentin, j. k. and williams, g. l. 1981. Fossil dinoflagellates: index to genera and species, 1981 edition. Bedford
Institute of Oceanography, Report Series, BI-R-8 1-12, 1-345.
lindemann, E. 1928. Abteilung Peridineae (Dinoflagellatae). 3-104. In engler, a. (ed.). Die naturlichen
Pflanzenfamilien, 2. Verlag von Wilhelm Engelmann, Leipzig.
loeblich, a. r. in 1969. The amphiesma or dinoflagellate cell covering. Proceedings of the North American
Paleontological Convention, G, 867-929.

loeblich, a. R., jr and loeblich, a. R. in 1966. Index to the genera, subgenera, and sections of the
Pyrrhophyta. Studies in Tropical Oceanography, 3, 1-94.
mathewes, r. w., heusser, L. E. and patterson, r. t. 1993. Evidence for a Younger Dryas-like cooling event
on the British Columbia coast. Geology , 21, 101-104.

matsuoka, k. and bujak, j. p. 1988. Cenozoic dinoflagellate cysts from the Navarin Basin, Norton Sound and
St. George Basin, Bering Sea. Bulletin of the Faculty of Liberal Arts, Nagasaki University, Natural Science,
29, 1-147.

matthiessen, j. 1991. Dinoflagellaten-Zysten im Spatquartar des europaischen Nordmeeres: Paldkologie und
Palao-oceanographie. Geomar Report, 1, 1-104.

morton, a. c. 1984. Coarse fraction of Plio-Pleistocene sediments from Deep Sea Drilling Project Hole 552A,
northeast Atlantic. Initial Reports of the Deep Sea Drilling Project, 81, 663-667.
mudie, p. j. 1992. Circum-arctic Quaternary and Neogene marine palynofloras: paleoecology and statistical
analysis. 347-390. In head, m. j. and wrenn, j. h. (eds). Neogene and Quaternary dinoflagellate cysts and
acritarchs. American Association of Stratigraphic Palynologists Foundation, Dallas, 438 pp.

- and short, s. k. 1985. Marine palynology of Baffin Bay. 263-308. In Andrews, j. t. (ed.). Quaternary
Studies of Baffin Island, West Greenland and Baffin Bay. George Allen and Unwin, London.

pascher, A. 1914. Uber Flagellaten und Algen. Bericht der Deutschen botanischen Gesellschaft, 36,
136-160.

PEACOCK, J. D., AUSTIN, W. E. N., SELBY, I., GRAHAM, D. K., HARLAND, R. and WILKINSON, I. P. 1992. Late
Devensian and Flandrian palaeoenvironmental changes on the Scottish continental shelf west of the Outer
Hebrides. Journal of Quaternary Science, 7, 145-161.

pisias, n. g., dauphin, j. p. and sancetta, c. 1973. Spectral analysis of Late Pleistocene-Holocene sediments.
Quaternary Research, 3, 3-9.

reid, p. c. 1974. Gonyaulacacean dinoflagellate cysts from the British Isles. Nova Hedwigia. 25, 579-637.
Roberts, D. G. and others. 1984. Initial reports of the deep sea drilling project, US Government Printing Office,
Washington, 923 pp.

HARLAND: NORTH ATLANTIC DINOFLAG ELLATES

281

ruddiman, w. f. 1987. Northern Oceans. 137-154. In ruddiman, w. f. and wright, h. e., jr. (eds). North
America and adjacent oceans during the last deglaciation. The Geology of North America, Geological Society
of America, Colorado, K-3, 501 pp.

— and glover, l. k. 1875. Subpolar North Atlantic circulation at 9300 yr bp: faunal evidence. Quaternary
Research. 5, 361-389.

— and McIntyre, a. 1981. The mode and mechanism of the last deglaciation: oceanic evidence. Quaternary
Research. 16, 125-134.

— and wright, h. e., jr. 1987. Introduction. 1-12. In ruddiman, w. f. and wright, h. e. (eds). North
America and adjacent oceans during the last deglaciation. The Geology of North America, Geological Society
of America, Colorado, K-3, 501 pp.

sarjeant, w. a. s. 1970. The genus Spiniferites Mantell, 1850 (Dinophyceae). Grana. 10, 74-78.
sejrup, h. p., jansen, e., erlenkeuser, h. and holtedahl, h. 1984. New faunal and isotopic evidence on the
late Weichselian-Holocene oceanographic changes in the Norwegian Sea. Quaternary Research, 21. 74-84.
selby, i. 1989. Quaternary Geology of the Hebridean continental margin. Unpublished Ph.D. Thesis,
University of Nottingham.

sernander, r. 1908. On the evidence of Postglacial changes of climate furnished by the peat-mosses of
Northern Europe. Geologiska Foreningen i Stockholm Forhandlingar, 30, 465-473.
shackleton, n. j. and hall, m. a. 1984. Oxygen and carbon isotope stratigraphy of Deep Sea Drilling Project
Hole 552A: Plio-Pleistocene glacial history. Initial Reports of the Deep Sea Drilling Project, 81. 599-609.
stoker, m. s., harland, r., morton, a. c. and graham, d. k. 1989. Late Quaternary stratigraphy of the
northern Rockall Trough and Faeroe-Shetland Channel, northeast Atlantic Ocean. Journal of Quaternary
Science, 4, 2 1 1-222.

stover, l. e. and evitt, w. r. 1978. Analyses of pre-Pleistocene organic-walled dinoflagellates. Stanford
University Publications, Geological Sciences, 15, 1-300.
turon, j.-l. 1978. Les dinoflagelles temoins des paleoenvironnements durant FHolocene dans l’Atlantique nord-
oriental. Signification paleohydrologique et paleoclimatique. Centre de Recherches de F Academie de Science
de Paris, Serie D, 286, 1861-1864.

— 1980. Dinoflagelles et environnement climatique. Les kystes de dinoflagelles dans les sediments Recents
de FAtlantique nord-oriental et leurs relations avec Fenvironnement oceanique. Application aux depots
Holocenes du Chenal de Rockall. Memoires du Museum nationale d'histoire naturelle, B27, 269-282.

— 1981. Le palynoplancton dans Fenvironnement actuel de FAtlantique nord-oriental. Evolution climatique
et hydrologique depuis le dernier maximum glaciaire. Thesis. Universite de Bordeaux , 698, 1-313.

veum, t., jansen, e., Arnold, m., beyer, i. and duplessy, j.-c. 1992. Water mass exchange between the North
Atlantic and Norwegian Sea during the past 28,000 years. Nature , 356, 783-785.
vernal, a. de, londeix, l., mudie, p. j., harland, r., morzadec-kerfourn, m.-t., turon, j.-l. and wrenn,
j. h. 1992. Quaternary organic-walled dinoflagellate cysts of the North Atlantic Ocean and adjacent seas:
ecostratigraphy and biostratigraphy. 289-328. In head, m. j. and wrenn, j. h. (eds). Neogene and Quaternary
dinoflagellate cysts and acritarchs. American Association of Stratigraphic Palynologists Foundation, Dallas,
438 pp.

wall, d. 1967. Fossil microplankton in deep-sea cores from the Caribbean Sea. Palaeontology, 10, 95-123.

— dale, b., lohmann, g. p. and smith, w. k. 1977. The environmental and climatic distribution of
dinoflagellate cysts in modern marine sediments from regions in the North and South Atlantic Oceans and
adjacent seas. Marine Micropaleontology, 2, 121-200.

wilson, g. j. 1973. Palynology of the middle Pleistocene Te Piki bed. Cape Runaway, New Zealand, New
Zealand J owned of Geology and Geophysics, 16, 345-354.

rex harland

Biostratigraphy and Sedimentology Group

Typescript received 24 May 1993 British Geological Survey, Keyworth

Revised typescript received 27 August 1993 Nottingham NG12 5GG, UK

282

PALAEONTOLOGY, VOLUME 37

APPENDIX I

List of samples

DSDP Hole 552A

DSDP Hole 610A

CSB No.

Depth (m)

CSB No.

Depth (m)

9139

010

10 1 16

0-12

9140

0-18

10 117

0-30

9141

0-30

10 118

040

9142

0-39

10 1 19

0-60

9143

0 51

10 120

0-75

9144

0-62

10 121

0-90

9145

0-71

10 122

106

9146

0-78

10 123

1-20

9147

0-89

10 124

1-35

9148

1-00

10 125

148

BGS Vibrocore 57/- 10/84

BGS Gravity Core 56/-10/47

CSB No.

Depth (m)

CSB No.

Depth (m)

9099

Sea-bed

9588

Sea-bed

9100

010

9589

010

9101

0-20

9590

0-20

9102

0-30

9591

0-30

9103

040

9592

0-40

9104

0-50

9593

0-50

9105

0-60

9594

0-60

9106

0-70

9595

0-70

9107

0-80

9596

0-80

9108

0-90

9597

0-90

9109

100

9598

1-00

91 10

1 10

9599

1 10

91 1 1

1-20

9600

1-20

91 12

1 30

9601

1 30

91 13

1-40

9602

1 40

9114

1-50

9603

1 50

9115

1 60

9604

1-60

9116

1-70

9605

1-70

9117

1-80

9606

1-80

9118

1-90

9607

1 90

9119

200

9608

2-00

APPENDIX 2

Taxonomic listing of the dinoflagellate cysts

The dinoflagellate cyst taxonomy for Quaternary dinoflagellate cysts consists of a mix of that derived from
palaeontology and phycology (see Harland 1982, 1983 for discussion). The list of taxa recovered in this study
reflects that mix but allows for the best circumscription of the tax pending the release of a new classification
of living and fossil dinoflagellates.

Division pyrrhophyta Pascher, 1914
Class dinophyceae Fritsch. 1929
Order peridiniales Haeckel, 1894
Family gonyaulacaceae Lindemann, 1928

Bitectatodinium tepikiense Wilson, 1973 (see Harland 1983, pi. 43, figs 3-4).

Impagidinium aculeatum (Wall) Lentin and Williams, 1981 (see Harland 1983, pi. 46, figs 1-3).

HARLAND: NORTH ATLANTIC DINOFLAGELL ATES

283

/. paradoxum (Wall) Stover and Evitt, 1978 (see Harland 1983, pi. 46, figs 4-5).

I. patulum (Wall) Stover and Evitt, 1978 (see Harland 1983, pi. 46, figs 6-7).

/. sphaericum (Wall) Lentin and Williams, 1981 (see Harland 1983, pi. 46, figs 8-9).

Lingulodinium machaerophorum (Deflandre and Cookson) Wall, 1967 (see Harland 1983, pi. 43, figs 5-6).
Nematosphaeropsis labyrinthus (Ostenfeld) Reid, 1974 (see Harland 1983, pi. 43, figs 7-8).

Operculodinium centrocarpum (Deflandre and Cookson) Wall, 1967 (see Harland 1983, pi. 43, figs 9 10).
Spiniferites elongatus Reid, 1974 (see Harland 1983, pi. 44, figs 7-8).

S. lazus Reid. 1974 (see Harland 1983, pi. 44, figs 1 1-12).

S. membranaceus (Rossignol) Sargeant, 1970 (see Harland 1983, pi. 45, figs 3-^4).

S. mirabilis (Rossignol) Sarjeant, 1970 (see Harland 1983, pi. 45, figs 1-2).

S. ramosus (Ehrenberg) Loeblich and Loeblich, 1966 (see Harland 1983, pi. 45, figs 5-6).

Family peridiniaceae Ehrenberg, 1832

Algidasphaeridium ? minutum (Harland and Reid) Matsuoka and Bujak, 1988 (see Harland 1992a, pi. 5.2, fig.
14).

Protoperidinium conicoides (Paulsen) Balech, 1974 (see Harland 1983, pi. 47, figs 2-3).

P. conicum (Gran) Balech, 1974 (see Harland 1983, pi. 47, figs 9-10).

P. leonis (Pavillard) Balech, 1974 (see Harland 1983, pi. 47, figs 7-8).

P. pentagonum (Gran) Balech, 1974 (see Harland 1983, pi. 48, figs 2-3).

P. subinerme (Paulsen) Loeblich IK, 1969 (see Harland 1983, pi. 47, figs 1 1-12).

Order gymnodiniales Lemmermann, 1910
Family polykrikaceae Kofoid and Swezy, 1921

Polykrikos schwartzii Butschli, 1873 (see Harland 1983, pi. 48, figs 10-12).

ICHNOFABRIC FROM THE UPPER JURASSIC
LITHOGRAPHIC LIMESTONE OF CERIN,
SOUTHEAST FRANCE

by C. GAILLARD, P. BERNIER, J. C. GALL, Y. GRUET, G. BARALE,

J. P. BOURSEAU, E. B U FFET A U T and S. WENZ

Abstract. The upper Kimmeridgian lithographic limestones of Cerin. France, are lagoonal deposits,
remarkable for the local occurrence of invertebrate burrows. Burrows are rare in the lower, well-laminated
lithographic limestones, attesting to the absence of autochtonous benthic animals and, therefore, to the
unfavourable life conditions on the lagoon floor. However, burrows are frequent in the upper lithographic
limestones, the most abundant being Tubularina lithographica. These burrows are small, partly filled by pellets
and were probably inhabited by intertidal polychaete worms. They are similar to Recent burrows restricted to
the intertidal area of the lagoon of Aldabra (Seychelles). T. lithographica is fossilized as an ‘open burrow',
attesting to the drying-out of the lagoon and the onset of lithification. Other trace fossils, dominated by
Thalassinoides and Rhizocorallium, are restricted to certain intermediate levels between the lower and upper
lithographic limestones. The resulting ichnoscquence may be the result of increasing periods of emersion in the
lagoon. Compared with other well known lithographic limestones (e.g. Solnhofen, Canjuers, Montsec), this
rich and unique ichnofabric clearly reflects the coastal location of the Cerin site.

Few of the well-known lithographic limestones contain trace fossils, as they were normally
deposited in environments (lacustrine, more or less deep marine environments - see Bernier and
Gaillard 1994) that were generally unfavourable to benthic life. For example, the intensively
sampled lithographic limestones from the Solnhofen area (Bavaria, Germany) have never yielded
invertebrate burrows except for a few tracks of dying animals such as crayfish and limulids (Barthel
et al. 1990). The Cerin lithographic limestone is unusual, however, in yielding a relatively rich
ichnofauna. This paper describes the most common type of trace fossil in this fauna, which is
referred to Tubularina lithographica igen. et isp. nov. The general sequence of ichnofabrics and their
environmental implications are also discussed.

MATERIAL AND METHODS

All of the studied samples were collected during a scientific excavation at Cerin, southern Jura
Mountains, Ain, France (Text-fig. 1). The excavation was made in a disused quarry, where the
lithographic limestone had been worked during the nineteenth century. The site had yielded a varied
range of well preserved animal and plant fossils, which have made the locality famous (Bourseau
et al. 1984). Since 1975, a bed-by-bed study of the whole formation has been organized. This
systematic investigation occurred in two areas, measuring 75 m2 (upper excavation) and 150 m2
(lower excavation) respectively, and produced much homogeneous palaeontological and sediment-
ological data. In particular, the study of the large upper bedding surfaces yielded many iclmologic
observations (Gaillard et al. 1991). The lithographic limestones are upper Kimmeridgian (Enay
et al. 1994). The palaeoenvironment corresponds to the margins of a shallow, tropical lagoon, lying
above a dead coral reef complex (Barale et al. 1985; Bernier et al. 1991).

(Palaeontology, Vol. 37, Part 2, 1994. pp. 285-304, 2 pls|

© The Palaeontological Association

286

PALAEONTOLOGY, VOLUME 37

The bioturbation of each bed was studied in the field. The most interesting specimens were
photographed and/or sampled. Many oriented and parallel sections (vertical and horizontal) were
made through bioturbated beds, making it possible to measure burrow orientation. Variations in
burrow density were detected by comparing corresponding surfaces from each bed. Some of the
slabs were polished to allow detailed observation and photography. Thin sections through selected
burrows were made. The pellets infilling the burrows and host lithographic limestone were observed
and compared by scanning microscopy.

SYSTEMATIC PALAEONTOLOGY
Ichnogenus TUBULARINA Gaillard, igen. nov.

Type species. Although ichnotaxa established at the genus-group level do not require a type species
(International Code of Zoological Nomenclature, 3rd Edition, 1985), it is suggested that Tubularina
lithographica isp. nov. be regarded as the effective type.

Derivation of name. From the tubular shape of the burrow.

Diagnosis. Small tubular open burrow with a sharply defined smooth wall, a few branches, and
frequent filling by well preserved faecal pellets.

Tubularina lithographica Gaillard, isp. nov.

Plate 1 ; Plate 2, figures 1-3; Text-figure 2

Holotype. Specimen number 286 300, FSL Collections of the Centre des Sciences de la Terre, University
of Lyon-1 (PI. 1 , fig. 3).

Type locality and horizon. Quarry at Cerin-Marchamp, Ain, France. Upper lithographic limestones, bed
number 274A (holotype in slab number 274A-W).

Derivation of name. From the lithographic nature of the host limestone.

Diagnosis. Small cylindrical burrow, a few centimetres long, up to 2 mm in diameter, with little

EXPLANATION OF PLATE 1

Figs 1-8. Tubularina lithographica Gaillard, igen. et isp. nov. Cerin, France; upper lithographic limestone
(upper Kimmeridgian). I, vertical section through bed 281, showing bioturbation limited to upper part of
bed; x 0 3. 2, vertical section through bed 354, showing the upper part only strongly bioturbated, and the
top of the bed outlined by a red-brown coloration and with prominent burrows; x 0-6. 3, holotype; vertical
section through bed 274A, showing a well developed vertical tunnel, with two branches, and an incomplete
filling by faecal pellets; the ichnofabric is complex, with two generations of burrows (see Text-fig. 9); x 0-6.
4, vertical section through bed 274A, showing specimen with complex branching; x F8. 5, vertical section
through bed 274A, showing specimen with incomplete filling by well sorted and arranged faecal pellets,
indicating the polarity; x4. 6-8, vertical sections through different examples of tunnels, showing variation
in the size, shape and distribution of the faecal pellet fill; 6, longitudinal section; x8; 7-8, transverse
sections ; x 1 2.

PLATE I

GAILLARD et al. , Jurassic ichnofabrics

PALAEONTOLOGY, VOLUME 37

branching. Tunnels sinuous and oriented in all directions relative to the bedding plane. Pellets filling
it ellipsoidal, micritic, without internal structure, and up to 1 mm in diameter.

Description

Burrow. The outline is generally very sharply defined (Text-fig. 2a). Transverse sections are circular (PI. 1, figs
7-8; Text-fig. 2b), while longitudinal ones do not display any significant increase in the diameter. Diameter
ranges from 0-5 mm to 2-0 mm. The total length is difficult to establish exactly but probably did not exceed
150 mm (the maximum depth of bioturbation in beds). Branching occurred but is rarely observed in section
(PI. 1, figs 3^4; Text-fig. 3). The burrow organization is rather complex. It penetrated the sediment more or
less sinuously, in very different directions, but mostly with a subvertical (70°-90°) or subhorizontal (0°-20°)
orientation (Text-fig. 4). From observations of vertical and horizontal parallel sections, the burrow system has
been reconstructed, as shown in Text-figure 5. The wall was smooth, without ornamentation, and generally
without lining. In some rare cases, thin section show a thin, dark lining which may have been detached (Text-
fig. 2b-c). This possibly resulted from a mucus coating. The burrow is filled rarely by micrite, but more
normally by sparite often including numerous pellets. Pellets do not line the burrow but clearly filled it.

Pellets. They are ellipsoidal, with a well preserved form (PL 1, figs 6-8; PL 2, figs 1-2; Text-fig. 2a-b). They
are generally in contact, but not crushed. Because their diameter is approximately half the diameter of the
containing burrow, no more than two rows of pellets could be observed in longitudinal section (PI. 1. figs 5-6)
and four specimens in transverse section (PL 1, figs 7-8). The pellets are micritic and exhibit the same
homogeneous ultrastructure as the enclosing limestone (PL 2, figs 3, 5). In some cases faecal pellets exhibit a
slight preferential dolomitization (Text-fig. 2a-b). The host rock is a very typical lithographic limestone
(Bernier 1994). It consists essentially of CaCO:! (99-5 per cent) and corresponds to a very pure and fine-
grained micrite. The grain size of the anhedral microcrystals is less than 4 /mi, frequently as small as 2 pm.

Comparisons. Granularia is not clearly defined. It was founded by Pomel (1849) on Algacites
granulatus Schlotheim, 1822, a species which Brongniart (1849) had also used as the basis for the
genus Phymatoderma ; Phymatoderma is therefore a synonym of Granularia. Both genera were

GAILLARD ET AL. : JURASSIC ICHNOFABRICS

289

text-fig. 2. Thin sections through Tubularina lithographic a ; Cerin, upper excavation, a, longitudinal section
of a typical burrow with sharp limits, sparitic filling, rounded micritic faecal pellets, and preferential
dolomitization of faecal pellets; x 10. b, transverse section with faecal pellets; x 12. c, transverse section of
burrow filled only with blocky calcite; an unusual detached dark lining is visible (also seen in b) which may

be an ancient mucus lining; x 12.

previously described as plant fossils, as were similar forms of Chondrites. According to the modern
interpretation summarized by Hantzschel (1975), Granularia corresponds to 'Elongated fillings of
burrows; long, diameter up to about 15 mm.; twig-shaped, with rather regular branching; walls
originally lined with clay particles...’. Granularia has not been frequently used in the literature.
Following the original description, Granularia looks a little like Tubularina. Compared with T.
lithographic a, ichnospecies that could be assigned to Granularia are of different size and more
frequently branched. Phymatoderma caelatum Saporta (1873, pi. 68, fig. 3), which is known from
the Upper Jurassic, is very probably a small burrow, 1 2 mm in diameter, filled with pellets. But
these are more elongated and irregular than in T. lithographic a. On the other hand, Granularia
repanda Pomel (see Saporta 1872, pi. 12, fig. la), whose rounded grains are probably small
ferruginous concretions, is very different. Granularia lumbricoides Heer is the most similar
ichnospecies but with more numerous, straighter, larger branches (Rothpletz 1896; Reis 1910).

When algal interpretations were abandoned, Granularia was often used for post-depositional
burrows in turbidites (Seilacher 1961; Ksiazkiewicz 1970, 1977; Crimes 1976; Leszczynski and
Seilacher 1991). These are very different from Tubularina. Indeed, Granularia occurs mainly more
or less horizontally on the sole of sandy layers (hypichnial ridges) and its diameter may be very
variable (Ksiazkiewicz 1970). Moreover, the specimens described by Seilacher (1961) are lined, not
filled, with mud pellets.

290

PALAEONTOLOGY, VOLUME 37

Coprulus oblongus described by Mayer (1952) corresponds to coprolites filling larger and more
complex burrows. Moreover, Coprulus must be restricted to a special kind of coprolite (Gaillard
1978). The closely related ichnogenus Tomaculum Groom, 1902, is a larger burrow (10-20 mm in
diameter) filled by larger elongate pellets (1-5 mm long, 0-5—1 -5 mm in diameter) and commonly
lying on bedding planes.

Tubularina resembles Trypanites Magdefrau, 1932, by the length of the tunnel, its sharply defined
edges, and its possible filling by the excrement of the producer. Trypanites is more or less straight,
however, usually vertical and unbranched. Moreover, Tubularina characterizes a firm ground (see
below) and Trypanites a hard ground.

Interpretation. Because of their very sharp contours in a rock which was formed from mud, and
because they are filled by sparry calcite, the burrows related to Tubularina must have been typically
‘open burrows’ in firm ground. Pellets are also very fragile grains which require special conditions
to be preserved. These taphonomic conditions may occur during emergent periods. The rapid
formation of firm grounds is easy because of the development of microbial mats at the surface of
the sediment (Gall et al. 1985; Bernier et at. 1991) but the formation of hard grounds is unlikely.

Worms or crustaceans may have been possible dwellers. Polychaete worms are more probable,
however, mainly because of the absence of scratch marks on the wall of the burrow, and the shape
and structure of pellets. Pellets from Tubularina are homogenous, without any structure, while
crustacean pellets commonly exhibit complex internal structures. Detailed comparison with modern
burrows in a similar environment provides further evidence for the polychaete hypothesis (see
below). Burrow-dwelling polychaete worms usually produce such ellipsoidal pellets. They are
formed by peristaltic movements in the intestine, coated with a thin mucus film, and then deposited
at the opening of the burrow, forming a small pile. The well-known species Heteromastus filiform is ,
which is very abundant in modern tidal flats, forms a small mound of faecal pellets at the surface
opening of its burrow (Schafer 1952, 1972). Normally, it is destroyed and the pellets, in spite of their
relative cohesion (Cadee 1979), are rolled and broken by currents. The penetration and preservation
of pellets in empty branches of the active burrow is possible. Alternatively, preferentially under
special conditions (e.g. during a long emergent period) worms die and the pellets remaining at the
surface may be preserved. Some of them can subsequently be introduced into the open, empty,
inactive burrow where they are well protected (Text-fig. 6). This seems the most probable hypothesis
for the genesis of the typical Tubularina lithographica specimens containing faecal pellets.

OTHER TRACE FOSSILS FROM THE CERIN LITHOGRAPHIC LIMESTONE

The following were the other main burrows found in the Cerin lithographic limestone, and occurred
only at a few levels. Others were very rare and may be related to well-known ichnogenera such as
Planolites or were indistinct traces.

EXPLANATION OF PLATE 2

Figs 1-3. Tubularina lithographica Gaillard, igen. et isp. nov. Cerin, France; upper lithographic limestone
(upper Kimmeridgian). 1, longitudinal thin section showing faecal pellets; x 20. 2, SEM view of micritic
faecal pellets and sparitic cement; x 60. 3, SEM view of detail of the micritic content of a faecal pellet;
x 1300.

Figs 4-6.- SEM views of Recent polychaete burrows; Dune-Jean- Louis creek shore, Aldabra. 4. longitudinal
section through burrow, showing faecal pellets and showing partly consolidated lime mud; x 15. 5, detail
of the faecal pellets from a similar burrow; x 50. 6, detail of the micritic content of a faecal pellet from the
burrow in 4; x 1300.

PLATE 2

GAILLARD et al Jurassic ichnofabrics

292

PALAEONTOLOGY, VOLUME 37

text-fig. 3. Tunnels of Tubularina lithographica in
vertical section, drawn from polished slabs, a-b, long
vertical tunnels (a — holotype). od, branching tun-
nels. e-g, long horizontal tunnels, h, burrow with
upper vertical tunnel and lower horizontal branched
tunnels, a, c-f, h, bed number 274A; b, bed number
306; g, bed number 296.

0 10 20 30 40 50 specimens bedding plane

text-fig. 4. Orientation of Tubularina
lithographica tunnels relative to the bed-
ding plane (studied from vertical polished
sections).

Thalassinoides Ehrenberg, 1944

This is a complex branching burrow with a horizontal network connected to the sediment-water
interface by vertical shafts. Specimens from Cerin were poorly preserved (Text-fig. 7a). Only the
horizontal network was visible with clear Y-shaped bifurcations and smooth walls. They are
probably related to Thalassinoides suevicus (Rieth, 1932). The trace is usually interpreted as a
feeding and dwelling burrow of a crustacean. Jurassic Thalassinoides containing the macrurous

GAILLARD ET AL. JURASSIC ICHNOFABRICS

293

crustacean Glyphaea have been described by Sellwood ( 1971 ). Glyphaea , which is known as a body
fossil from the Cerin lithographic limestones, is a possible excavator of Thalassinoides .

Rhizocorallium Zenker, 1836

This is a U-shaped spreite-burrow parallel to the bedding plane. Specimens were rare at Cerin and
not well preserved (Text-fig. 7b). They resembled slightly burrows from the Upper Jurassic (Ftirsich
1974r/) and Lower Cretaceous (Basan and Scott 1979) that are related to Rhizocorallium irregulare
Mayer, 1954. This trace is usually interpreted as the burrow of a deposit-feeder, probably a
crustacean (Ftirsich 1 974<r/, 1974/i).

294

PALAEONTOLOGY, VOLUME 37

text-fig. 7. Trace fossils from the lithographic limestones of Cerin. a, Thalassinoides ; x 0T6. b, Rhizocor allium ;

x 030.

COMPARISON WITH MODERN ENVIRONMENTS

The wide distribution of Upper Jurassic lithographic limestones has no direct equivalent in the
Recent. The Aldabra Lagoon (Indian Ocean, Seychelles) is probably the closest modern
environment to the Cerin Lagoon (Gaillard et al. 1991, 1994). The lagoon of the Aldabra atoll
has a surface area of 200 km2, and is only a few metres deep. It communicates with the open sea
through several channels. Twice a day, it is filled and emptied following a tidal rhythm. Calcareous
material from benthic production and bioerosion is permanently reworked and distributed in the
lagoon area according to the grain size. Consequently, a very pure lime mud is deposited along the
most protected lagoon shore lines. It is frequently covered by microbial films which act as sediment
stabilizers (Text-fig. 8a). These mud flats emerge for various periods of time, according to the
distance from the shore. They are actively burrowed by different organisms, mainly crustaceans
(Farrow 1971). The Jurassic environment of Cerin was certainly very different, principally because
Aldabra is an atoll lagoon communicating easily with the open sea. Mud flats in the Aldabra lagoon
are located in the most protected areas of the lagoon. The Cerin lithographic limestone was
deposited in a very restricted and hostile lagoonal environment, located within an extensive tract of
land. This explains the rarer benthic fauna inhabiting the Cerin lagoon. Nevertheless, there are
similarities, especially in the creeks of the southern edge of the Aldabra lagoon, for example in the
Dune-Jean-Louis Creek (Gaillard et al. 1994).

Mudflats are located along the protected southern shore of the Aldabra lagoon and are heavily
burrowed by the crustacean Alpheus crassissimus. These burrows, which look like Thalassinoides,
are situated near extreme low water and are covered at neap high tides. They are formed in a typical
soft ground. In some more littoral and sheltered areas, mud forms a firm ground which is also
intensively bioturbated. This firm ground is the result of considerable dewatering and efficient
binding by microbial mats built by Microcoleus chtonoplastes (Text-fig. 8a-b). Different kinds of
burrows are inhabited by worms and crustaceans (Text-fig. 8c). The most abundant burrows
strongly resemble Tubularina lithographica because of their length and their filling by spheroidal
faecal pellets (PI. 2, figs 4-6). Living burrowing polychaete worms, belonging to Eunicidae
( Marphysa mossamhica) and to Nereidae ( Perinereis cultrifera ), have been found in this environment
(determined by Y. Gruet and P. Gillet).

Similar environmental conditions may have occurred in the most coastal areas of the Cerin
lagoon. In the centre of the lagoon, poisonous anoxic water hindered normal benthic life.
Exceptionally some infaunal decapods colonized the lagoon floor (forming Thalassinoides ) when
sufficiently oxygenated. The Thalassinoides community may be compared to the recent Alphaeus
community of the lagoon of Aldabra. The burrowing shrimp Alphaeus is common in fine-grained
lime sediments, for example throughout the Florida Keys and the Bahamas (Shinn 1968). It is most

GAILLARD ET AL. JURASSIC ICHNOFABRICS

295

text-fig. 8. a, calcareous mud flat (‘blanc d'Espagne’) at low tide, partly covered by eroded Microcoleus mat;
Abbott’s creek, Aldabra. b, close-up of eroded Microcoleus mat; Dune-Jean-Louis creek, Aldabra. c, partly
consolidated (water loss) lime mud showing, in section, many burrows sometimes including faecal pellets
(arrow); Dune-Jean-Louis creek shore, area with long emersive periods, Aldabra.

abundant in the intertidal zone where it digs complex burrows of Thalassinoides- type. In the Cerin
lagoon, generally only microbial mats flourished. If emergent for too long, they would be
progressively altered by mechanical and biological processes. Such a relationship between microbial
development and animal benthic life can be observed in many tidal environments. Gerdes and
Krumbein (1986) have described the relationship between microbial mats and infaunal invertebrates
in a supratidal depositional environment from the North Sea, which is a well oxygenated open sea.
The texture, composition and degradation products of the microbial mats, which are dominantly
formed by Microcoleus chtonoplastes in the most sheltered areas, are observed to control the
abundance of burrowing animals. The infaunal population is low because of anaerobic conditions
under the mat surface, where ammonia and sulfide concentrations are relatively high. One of the

296

PALAEONTOLOGY, VOLUME 37

most abundant burrowing organisms is the polychaete worm Pygospio elegans which is able to
recolonize rapidly the fresh sediment, thanks to variable reproduction cycles and a high
reproduction rate. In comparison, the burrower corresponding to Tubularina lithographica was
probably also an opportunistic organism.

In the Cerin lagoon, greater emersion probably caused the mechanical degradation and the
weathering of microbial mats, leading to mud colonization by burrowers. Indeed, desiccation cracks
occur at the surface of some beds, mainly in the transition levels of lithographic limestones,
although only rarely in the lower (probably because of short emersions) and upper lithographic
limestones (probably because often destroyed by bioturbation). Many cases of microbial mats
partly destroyed by burrowing animals can be observed in the Recent. For example, in the most
sheltered parts of the Aldabra lagoon, the degradation of the Microcoleus chtonoplastes mats by
burrowers frequently begins along desiccation cracks (Text-fig. 8b).

RESULTING ICHNOFABRIC

The Tubularina lithographica ichnofabric was examined in several oriented bed sections, where
burrow density and orientation could be studied. The most obvious observation was that
bioturbation was more frequent in the upper part of beds (PI. 1, figs 1-2). The lower part of beds
generally had no or very few burrows. The most complete ichnofabric showed no more than two
kinds of burrow (Text-fig. 9). First, indistinct irregular burrows filled by ambiant micrite occurred.

text-fig. 9. Examples of complex ichnofabric in the
upper lithographic limestones of Cerin. Bioturbation
by large indistinct burrows filled by light micrite (b),
and Tubularina lithographica (t). The first are cross-
cut by the second (X). On the right, the horizontal top
of the initial deposit of thinner dark micrite is visible.
It is probably consolidated by a thin microbial mat
(m); natural size.

These early burrows excavated the superficial microbial mat, when visible, and had been
subsequently cut through by Tubularina. This indicates that Tubularina was a late burrow made in
a firm ground. The maximum density of Tubularina was always a few centimetres below the top bed

GAILLARD ET AL. \ JURASSIC ICHNOFABRICS

297

text-fig. 10. Density of burrows at different depths, in beds of Cerin lithographic limestone of different
thicknesses, showing the development of bimodal distributions in the thicker beds.

text-fig. 11. a, irregular top surface of Cerin lithographic limestone beds bioturbated by Tubularina
lithographic a, showing late selective erosion, probably by dissolution; xO 1. B, detailed view of burrows; x I

surface, probably reflecting the deep development of the horizontal galleries of the burrow system
(Text-fig. 10). This depth of maximum burrow density increased with bed thickness (Text-fig.
10a-c). The more that available fresh sediment was abundant, the more the burrows were developed
and deeply settled. Curiously, in very thick beds (more than 120 mm in the case of the Cerin
lithographic limestones), two maxima of bioturbation sometimes occurred (Text-fig. 10d-e), the
upper one being the most important. This possibly illustrates two successive stages of sedimentation

298

PALAEONTOLOGY, VOLUME 37

in a relatively short time. Slight differences in the nature or consistency of the sediment could have
controlled the vertical development of the burrow system. In some cases the two stages of
sedimentation were proved when the lower part of the bed was topped by a slight microbial mat.

Observation of Tubularina lithographica also helped to understand the late evolution of the
lithographic limestones. Only in the upper part of the formation, where T. lithographica was
abundant (see below), the bed tops exhibit an irregular red-brown mamillate surface. Each knoll
was often bristling with prominent Tubularina (PI. 1, fig. 2; Text-fig. 11). These millimetre-scale
reliefs corresponded primarily to holes, then to their late sparitic filling, and after to the selective
ablation of the host sediment. Selective erosion, perhaps by dissolution, was possible during late
diagenesis. This phenomenon could have amplified the slight initial difference between the two
lithofacies of the Cerin quarry (Bausch et al. 1994). These (‘flinze’ and 'faulen’) differ in their
content of insolubles.

SEQUENCE OF TRACE FOSSILS AND GENERAL INTERPRETATION

Tubularina lithographica was the most abundant burrow found in the Cerin lithographic limestone.
It occurred at numerous levels, but only in the upper limestones (beds number 255 to 404 from the
scientific excavation; Text-fig. 12). As shown earlier, T. lithographica clearly indicates long periods
of emergence and a very cohesive sediment.

Rhizo cor allium occurred only in rare beds (beds number 250C to 270) in the lower part of the
vertical range of T. lithographica. In bed number 270, Rhizocor allium was associated with
Thalassinoides. The association of Rhizo cor allium irregulare and Thalassinoides suevicus is well-
known in shallow Jurassic environments (Fiirsich 1974a; Heinberg and Brikelung 1984), but
according to Fiirsich (1974a) the two trace-makers were probably not contemporaneous. The
bioturbation by Rhizocor allium took place at a later stage than that of Thalassinoides , when the
sediment became sufficiently cohesive.

Thalassinoides was also found in only a few beds, although more frequently than Rhizo cor allium .
It appeared lower, at the base of the lithographic limestones, and disappeared at the same level as
Rhizocor allium.

Three levels could therefore be distinguished: (1) a lower-level without traces or with very rare
Thalassinoides , (2) a short transition-level with Thalassinoides and Rhizocor allium, and (3) an upper-
level with abundant Tubularina lithographica. This ichnologic succession characterizes a very precise
shallowing upward sequence. The gradual increase in periods of emersion induced changes in
burrowing communities, probably controlled strongly by the increase in cohesion of the substrate.
This environmental interpretation is supported by other palaeontological data (Text-fig. 12). The
lower level showed dominant marine influence, with the introduction into the lagoon of open-sea
organisms. The fauna was characterized by the presence of ammonites and shrimps, and the
dominance of fishes and algae. In contrast, the upper level showed an increase in continental
influence, with the disappearance (ammonites) or reduction in abundance (fishes) of marine
animals, and the dominance of terrestrial plants ( Zamites ). Microbial mats were also deformed
(Bernier et al. 1991) but well preserved in the lower level, where bioturbation was limited or absent.
They were invisible or poorly preserved in the upper level probably because it was highly
bioturbated (Text-fig. 12).

Detailed ichnological sequences have been previously described from Jurassic shallow marine
environments. For example, Farrow (1966) has shown a bathymetric zonation of trace fossils in
Yorkshire Jurassic sections, where annelid burrows always precede crustacean burrows ( Rhizo -
corallium, Thalassinoides ) in the sequence, indicating a slight deepening of the water. At Cerin, the
sequence culminated with the reappearance of special annelid burrows (Tubularina) in more coastal
and restricted environments.

In the Upper Jurassic sediments from Boulonnais (northern France), Ager and Wallace (1970)
have shown the following clear sequence, reflecting shallowing and emergence: (1) horizontal

GAILLARD ET AL. : JURASSIC ICHNOFABRICS

299

text-fig. 12. Stratigraphical distribution of trace fossils and body fossils in the lithographic limestones of
Cerin, together with an interpretation of the environment.

300

PALAEONTOLOGY, VOLUME 37

Rhizocorallium, (2) large Thalassinoides , (3) obliquely oriented Rhizocor allium, and (4) Diplo-
craterion. Some elements of this sequence occur at Cerin (horizontal Rhizocorallium and
Thalassinoides ), but the last part is different. The replacement of the obliquely oriented
Rhizocorallium and Diplocraterion communities by Tubularina lithographica , probably reflects the
more restricted lagoonal conditions.

As T. lithographica seemed to be an excellent paleobathymetric index for marginal environments,
its occurrence and abundance were compared from two nearby sections in the Cerin lithographic
limestones (Text-fig. 13). The first (Section A) was in an area where lithographic limestones are

BEDS SECTIONS

A B

■ no data

BEDS

BIOTURBATED

BEDS

/\ /\

39%

53%

f3l Portlandian limestones

f?l Lithographic limestones

IT1 Coralllan and perl-coralllan substrate

text-fig. 13. Vertical and horizontal variation of Tubularina lithographica bioturbation in the upper

lithographic limestones in the Cerin lagoon.

thick. The second (Section B) was 50 m away and corresponded to the scientific excavation where
lithographic limestones are thinner and near their rocky substrate (Bernier et al. 1994). In the two
sections, the ichnofabric with T. lithographica appeared at the same level (bed number 255) and only
became abundant in the upper part. Four main intervals with increasing density of bioturbation

GAILLARD ET A L JURASSIC ICHNOFABRICS

301

were distinguished (SI to S4) and interpreted as shallowing upward sequences. It was interesting,
however, to note that the T. lithographica ichnofabric was visible only in thirty nine per cent of beds
from Section A (Text-fig. 14) but in fifty three per cent of beds from section B. This probably

text-fig. 14. Detail of ichnologic sequence S3 from section A (see Text-fig. 13), showing beds 363 to 368;

xO-25.

indicates that Section A was situated nearer the centre of the lagoon, which was deeper with
probably less frequent and shorter emersions than at the edge. ,It confirmed the very coastal
situation of the scientific excavation area, where many trackways of terrestrial animals have been
observed (Bernier et al. 1982, 1984; Gaillard el al. 1991 ).

CONCLUSIONS

A detailed bed by bed investigation of the upper Kimmeridgian lithographic limestones of Cerin has
revealed abundant but poorly diversified invertebrate trace fossils. This underlines the distinctive
nature of Cerin compared with other well known lithographic limestones sites. The most abundant
trace fossils ( Tubularina lithographica Gaillard) are small, branched burrows, generally filled by
spheroidal faecal pellets and sparite, indicating an emerged firm substrate. It was probably
produced by a polychaete worm similar to Recent ones living in intertidal environments like the
Aldabra lagoon (Seychelles). Other trace fossils, mainly Thalassinoides and Rhizocor allium, are rare
and indicate ‘deeper ’-water conditions. The Cerin trace fossils were organized in the following
shallowing upward ichnologic sequence: Thalassinoides - Rhizocor allium Tubularina. The

T. lithographica ichnofabric is probably a very precise palaeoecologic index for marginal marine
environments corresponding to restricted shallow lagoons.

Acknowledgements . The field and laboratory research relating to the scientific excavation of Cerin is mainly
funded by the CNRS-URA 1 1. Other financial support and equipment were provided by the Conseil General
de l’Ain, the Musee Guimet d’Histoire Naturelle de Lyon, and the Museum de Geneve. The French firms
Fleury-Michon and Rhone-Poulenc also sponsored this work. The expedition to the atoll of Aldabra was
sponsored by the Ministere de la Cooperation, the Seychelles Islands Foundation and BBS Production.
Scanning electron micrographs were made in the CMEABG of the University Lyon- I. Authors are grateful

302

PALAEONTOLOGY, VOLUME 37

to P. B. Wignall and an anonymous reviewer for reading and commenting on the manuscript. We would also
like to acknowledge the technical contribution of A. Armand, N. Podevigne, G. Sirven and J. C. Reniaud.

REFERENCES

ager. d. v. and Wallace, p. 1970. The distribution and significance of trace fossils in the uppermost Jurassic
rocks of the Boulonnais, northern France. 1-18. In crimes, t. p. and harper, j. c. (cds). Trace fossils.
Geological Journal, Special Issue , 3, 547 pp.

BARALE, G., BERNIER, P., BOURSEAU, J.-P., BUFFETAUT, E., GAILLARD, C., GALL, J.-C. and WENZ, S. 1985. Coin, line
lagune tropicale au temps des dinosaures. Centre National de la Recherche Scientifique, Paris, et Museum de
Lyon, Lyon, 136,pp.

barthel, K. w., Swinburne, N. H. M. and Conway morris, s. 1990. Solnhofen , a study in Mesozoic palaeontology.
Cambridge University Press, Cambridge, 236 pp.

basan. p. b. and scott, r. w. 1979. Morphology of Rhizocorallium and associated traces from the lower
Cretaceous Purgatoire Formation, Colorado. Palaeogeography , Palaeoclimatology, Palaeoecology , 28, 5-23.
BAUSC’H, W. M., VIOHL, G., BERNIER, P., BARALE, G., BOURSEAU, J.-P., BUFFETAUT, E., GAILLARD, C., GALL, J.-C. and

wenz, s. 1994. Eichstatt and Cerin: geochemical comparison and definition of two different plattenkalk
types. 107-125. In bernier, p. and gaillard, c. (eds). Calcaire lithographique, sedimentologie, paleontologie,
taphonomie. Geobios, Memoire Special , 16, 348 pp.

bernier, p. 1994. For a reinstatement of “lithographic” a precise word to define a precise limestone. 307-312.
In bernier, p. and gaillard, c. (eds). Calcaire lithographique, sedimentologie, paleontologie, taphonomie.
Geobios, Memoire Special, 16, 348 pp.

BARALE, G., BOURSEAU, J.-P., BUFFETAUT, E., DEMATHIEU, G., GAILLARD, C., GALL, J.-C. and WENZ, S. 1982.
Trace nouvelle de locomotion de Chelonien et figures sedimentaires associees dans les calcaires
lithographiques de Cerin (Kimmeridgien superieur, Ain, France). Geobios, 14. 447-467.

- 1 984. Decouverte de pistes de Dinosaures sauteurs dans les Calcaires
lithographiques de Cerin (Kimmeridgien superieur, Ain, France). Geobios, Memoire Special , 8, 177-185.

gaillard, c., gall, J.-C. and wenz, s. 1991. The paleoecological excavations at Cerin
(Southern French Jura Mountains), results and interpretation. Table Ronde Internationale ‘ Calcaire
lithographique' , Lyon, 8-10 Juillet 1991, Guidebook, 35 pp.

and gaillard, c. (eds) 1994. Calcaire lithographique, sedimentologie, paleontologie, taphonomie.
Table Ronde Internationale ‘Calcaire lithographique’. Geobios, Memoire Special , 16. 348 pp.

barale, G., bourseau, J.-P., buffetaut, E., gall, j.-c. and wenz, s. 1994. The underlying
substrate of the Cerin lithographic limestone. 13-24. In bernier, p. and gaillard, c. (eds). Calcaire
lithographique, sedimentologie, paleontologie, taphonomie. Geobios, Memoir Special, 16, 348 pp.

— gall, j.-c., barale, g., bourseau, j.-p., buffetaut, e. and wenz, s. 1991 . Morphogenetic impact of
microbial mats on surface structures of lithographic limestones (Cerin area, Kimmeridgian, France).
Sedimentology, 38, 127-136.

bourseau, j.-p., buffetaut, e., bernier, p., gaillard, c., gall, j.-c., barale, g. and wenz, s. 1984. Fa carriere
des calcaires lithographiques de Cerin (Ain, commune de Marchamp). Vie et extinction d’une exploitation
communale sur un gisement paleontologique celebre. Nouvelles Archives du Museum cT Histoire Naturelle de
Lyon, 22, 21-30.

brongniart, a. t. 1849. Tableau des genres de vegetaux fossiles consideres sous le point de vue de leur
classification botanique et de leur distribution geologique. Dictionnaire Universel d' Histoire Naturelle, 13,
1-127.

cadee, g. c. 1979. Sediment reworking by the polychaete Heteromastus filiformis on a tidal flat in the Dutch
Wadden Sea. Netherlands Journal of Sea Research, 13, 441^156.
crimes, t. p. 1976. Trace fossils of an Eocene deep-sea sand fan, northern Spain. 71 -90. In crimes, t. p. and
harper, j. c. (eds). Trace fossils. Geological Journal Special Issue, 3, 351 pp.
ehrenberg, k. 1944. Ergiinzende Bemerkungen zu den Seinerzeit aus dem Miozan von Burgschleinitz
beschriebenen Gangkernen und Bauten dekapoder Krebse. Paldontologischc Zeitschrift, 23, 354-359.
ENAY. R,. BERNIER, P., BARALE, G., BOURSEAU, J.-P., BUFFETAUT, E., GAILLARD, C., GALL, J.-C. and WENZ, S. 1994.
Fes ammonites des calcaires lithographiques de Cerin (Ain, France): stratigraphies et taphonomie. 25-36.
In bernier, p. and gaillard, c. (eds). Calcaire lithographique, sedimentologie, paleontologie, taphonomie.
Geobios, Memoire Special, 16, 348 pp.

farrow, cl E. 1966. Bathymetric zonation of Jurassic trace fossils from the coast of Yorkshire, England.
Palaeogeography, Palaeoclimatology, Palaeoecology, 2, 103-151.

GAILLARD ET AL. JURASSIC ICHNOFABRICS

303

— 1971. Back-reef and lagoonal environments of Aldabra atoll distinguished by their crustacean burrows.
Zoological Society of London , Symposium, 28, 455-500.

fursich, F. T. 1974«. Corallian (Upper Jurassic) trace fossils from England and Normandy. Stuttgarter
Beitrdge zur Naturkunde aits dem Staatlichen Museum fur Naturkunde in Stuttgart, Serie B, 13, 1-52.
1974/?. Ichnogenus Rhizocor allium. Palaontologische Zeitschrift, 48, 16-28.
gaillard, c. 1978. Revision de I’ichnogenre Coprulus Richter & Richter, 1939, et description de quelques
nouvelles especes du Jurassique superieur. Geobios, 11, 439 — 455.

— BERNIER, P., GALL, J.-C., BARALE, G., BOURSEAU, J.-P., BUFFETAUT, E. and WENZ, S. 1991. IchnofacieS of
lithographic limestones of Cerin and their environmental significance. Table Ronde Internationale L Calc air e
lithographique ’, Lyon, 8-10 Juillet 1991 , Abstracts, 23.

— and gruet, y. 1991. Le lagon d’Aldabra (Seychelles), un modele pour le paleomilieu de Cerin
(Kimmeridgien superieur, France). Table Ronde Internationale ' Calc air e lithographique', Lyon, 8-10 Juillet
1991 , Abstracts, 22.

- 1994. Le lagon d’Aldabra (Seychelles, Ocean Indien), un modele pour le paleoenvironnement
de Cerin (Kimmeridgien sup.. Jura meridional, France). 331 348. In bernier, p. and gaillard, c. (eds).
Calcaire lithographique, sedimentologie, paleontologie, taphonomie. Geobios, Memoire Special, 16, 348 pp.
GALL, J.-C., BERNIER, P., GAILLARD, C., BARALE, G., BOURSEAU, J.-P., BUFFETAUT, E. and WENZ, S. 1985. Influence
du developpement d’un voile algaire sur la sedimentation et la taphonomie des calcaires lithographiques.
Exemple du gisement de Cerin (Kimmeridgien superieur. Jura meridional franqais). Comptes Rendus de
T Academic des Sciences, Paris , Serie 2, 301, 547-552.

gerdes, G. and krumbein, w. e. 1986. Potentielle Silikoklastische Stromatolithe des unteren Supralitorals
(siidliche Nordsee) als begrenzende Faktoren gang- und rohrenbauender mariner Evertebraten. Neues
Jahrbuch fur Geologie und Paldontologie, Abhandlungen, 172. 163- 191.

GROOM, T. 1902. The sequence of the Cambrian and associated beds of the Malvern Hills. Quarterly Journal
of the Geological Society , London, 58, 89-149.

hantzschell, w. 1975. Trace fossils and problematica. 1-269. In moore, r. c. (eds). Treatise on invertebrate
paleontology. Part W, Miscellanea, Supplement 1, Geological Society of America and University of Kansas
Press, Boulder, Colorado, and Lawrence, Kansas, 269 pp.
ueinberg, c. and brikelund, T. 1984. Trace-fossil assemblages and basin evolution of the Vardekloft
formation (Middle Jurassic, Central East Greenland). Journal of Paleontology, 48, 362-397.
ksiazkiewicz, m. 1970. Observations on the ichnofauna of the Polish Carpathians. 283-322. In crimes, t. p.
and harper, j. c. (eds). Trace fossils. Geological Journal Special Issue, 3, 547 pp.

— 1977. Trace fossils in the Flysch of the Polish Carpathians. Palaeontologia Polonica, 36, 1-208.
leszczynski, s. and seilacher, a. 1991. Ichnocoenoses of a turbidite sole. Ichnos , 1, 293-303.
magdefrau, k. 1932. Uber einige Bohrgange aus dem unteren Muschelkalk von Jena. Palaontologische

Zeitschrift, 14, 150-160.

mayer, g. 1952. Neue Lebensspurren aus dem Unteren Hauptmuschelkalk (Trochitenkalk) von Wiesloch:
Coprulus oblongus n. sp. und Coprulus sphaeroideus n. sp. Neues Jahrbuch fur Geologie und Paldontologie,
Monatshefte, 8, 376-379.

pomel, a, 1849. Materiaux pour servir a la flore fossile des terrains jurassiques de la France. Amtlicher Bericht
liber die Versammlung Deutscher Naturforscher und Aerzte, 25 (September 1847), 332-354.
reis, O. M. 1910. Zur Fucoidenfrage. Jahrbuch der Kaiserlich-Koniglichen Geologischen Reichsanstalt, Wein , 59,
615-638.

rothpletz, a. 1896. Ueber die Flysch-Fucoiden und einige andere fossile Algen, sowie fiber liasische,
diatomeen ffihrendc Hornschwamme. Zeitschrift der Deutschen Geologischen Gesellschaft, 48, 854-914.
SAPORTA, G. de. 1872-1873. Paleontologie frangaise, plant es jurassiques. Tome 1: Introduction, Algues,
Equisetacees, Characees, Fougeres. Masson ed., Paris, 506 pp.
schafer, w. 1952. Biogene sedimentation im Gefolge von Bioturbation. Senckenbergiana, 33, 235-246.

— 1972. Ecology and paleoecology of marine environments. University of Chicago Press, Chicago, 568 pp.
schlotheim, e. f. von 1822. Ncichtrdge zur Petrefactenkunde. Gotha, 100 pp.

seilacher, a. 1961. Paleontological studies on turbidite sedimentation and erosion. Journal of Geology, 70,
227-234.

sellwood, b. w. 1971. A Thalassinoides burrow containing the crustacean Glyphaea udressieri (Meyer) from
the Bathonian of Oxfordshire. Palaeontology, 14, 589-591.
shinn, E. a. 1968. Burrowing in recent lime sediment of Florida and the Bahamas. Journal of Paleontology , 42,
879-894.

304

PALAEONTOLOGY, VOLUME 37

zenker, J. c. 1836. Historisch-topographisches Taschenbuch von Jena unci seiner Umgebung besonders in
naturwissenschaftlicher und medicinischer Beziehung. Jena, 338 pp.

C. GAILLARD
P. BERNIER

Universite Claude- Bernard, Lyon-1
Centre des Sciences de la Terre
Unite de Recherche Associee au C.N.R.S. 1 I
43 boulevard du 11 Novembre 1918
69622 Villeurbanne Cedex, France

J.-C. GALL

Universite Louis Pasteur
Institut de Geologie
1 rue Blessig

67084 Strasbourg, France
Y. GRUET

Universite de Nantes
Laboratoire de Biologie Marine
2 rue de la Houssiniere
44000 Nantes, France

G. BARALE

Universite Claude-Bernard, Lyon-1
Laboratoire de Paleobotanique
Unite de Recherche Associee au C.N.R.S. 11
43 boulevard du 11 Novembre 1918
69622 Villeurbanne Cedex, France

J.-P. BOURSEAU

Universite Claude-Bernard, Lyon-1
Centre des Sciences de la Terre
43 boulevard du II Novembre 1918
69622 Villeurbanne Cedex, France

E. BUFFETAUT

Universite Pierre et Marie Curie
Laboratoire de Paleontologie des Vertebres
Unite de Recherche Associee au C.N.R.S. 1433
4 place Jussieu

75252 Paris Cedex 05, France

S. WENZ

Typescript received 2 July 1993

Revised typescript received 24 January 1994

Museum National d'Histoire Naturelle
Laboratoire de Paleontologie
8 rue de Buffon
75005 Paris, France

THE ICHNOGENUS BEACON 1TES AND ITS
DISTINCTION FROM ANCORICHNUS AND

TAENIDIUM

by D. G. KEIGHLEY and R. K. PICKERILL

Abstract. Beaconites is a trace-fossil name that has been adopted indiscriminantly for unlined, lined,
unwalled, thinly walled, and thickly walled, meniscate backfilled burrows. The confusion is further exacerbated
by the inconsistent use of the terms 'wall' and ‘lining’. A wall and a lining (a type of wall) are herein restricted
to features actively constructed by the burrower, and are considered distinct from peripheral features produced
by simple excavation or during locomotion. Differences in the type of meniscate backfilling are also recognized,
and may assist in the distinction of ichnotaxa. Beaconites , and likewise the type ichnospecies B. antarcticus , is
a lined (walled) meniscate trace fossil; B. barret ti , the ichnospecies most popularly assigned to the ichnotaxon,
is actually unlined and unwalled, and cannot therefore be included within Beaconites. Recent emendments to
Taenidium describe it essentially as an unlined meniscate backfilled burrow. The diagnosis of Taenidium is,
however, further emended to clarify that it is an unwalled structure. Forms previously assigned to B. barretti
can therefore be included within Taenidium as T. barretti. Emendments to the original diagnosis of Ancorichnus
describe this trace fossil as a walled ichnotaxon. These emendments are rejected because this would place the
ichnogenus in junior synonymy with Beaconites ; instead, the original diagnosis of Ancorichnus is re-established.
Two ichnospecies, A. capronus and A. coronus are, nevertheless, considered to be separate ichnospecies of
Beaconites , namely B. capronus and B. coronus. The type ichnospecies, A. ancorichnus , is distinguished by a
structured mantle peripheral to a meniscate core. The mantle is not considered as a wall structure since it is
formed by the locomotive behaviour of the burrow producer.

Several authors have commented upon the ichnotaxonomic problems associated with meniscate
trace fossils. For example, Frey et al. (1984) discussed Ancorichnus Heinberg, 1974, and Scoyenia
White, 1929; D’Alessandro and Bromley (1987) discussed both Muensteria Sternberg, 1833, and
Taenidium Heer, 1877. These authors, with, for example. Squires and Advocate (1984), O’Sullivan
et al. (1986), Briick (1987) and Gordon (1988) also suggested that further synonyms may exist for
these and other meniscate trace fossils, including Beaconites Vialov, 1962.

One of the major problems with addressing the potential synonymies of Beaconites is the nature
of this ichnotaxon’s margins, its internal structure and, to a lesser extent, its overall size and
orientation. Such diagnostic criteria (‘ichnotaxobases’; Bromley 1990) have been confused and
applied indiscriminately within Beaconites. Accordingly, Beaconites has been variably described as
a small, walled meniscate burrow (Vialov 1962), a large welted meniscate burrow (Hantzschel 1975),
a smoothly-lined meniscate burrow (Bradshaw 1981), an unlined meniscate burrow (Frey et al.
1984; Squires and Advocate 1984; Briick et al. 1985), and a weakly or unwalled meniscate burrow
(D’Alessandro and Bromley 1987).

Since many of the terms employed in distinguishing Beaconites and related ichnotaxa have thus
been used with different or occasionally (and more problematically) unspecified meanings, the
terminology adopted for this contribution is initially summarized. The diagnostic criteria of
Beaconites are then re-examined, particularly regarding its distinction from the morphologically
similar ichnotaxa Ancorichnus and Taenidium. This is followed by revised synonymy listings for
Ancorichnus. Beaconites , and Taenidium.

| Palaeontology, Vol. 37, Part 2, 1994, pp. 305-337, 1 pl.|

© The Palaeontological Association

306

PALAEONTOLOGY, VOLUME 37

TERMINOLOGY

Backfill structure ( backfilled burrow)

Bioturbation in softgrounds caused by the active redeposition of sediment (active fill) immediately
behind a burrowing animal. Essentially, it is produced by the axial migration of a burrow that itself
is not preserved (Bromley 1990). In contrast, an open burrow is occupied and maintained by an
animal and passively (or occasionally actively) filled later (Bromley 1990, p. 266). Backfilling
probably assists in the organism’s forward movement (Chamberlain 1 9 7 1 ; Heinberg 1974) and the
transport of such material to the rear of the producer may occur either externally, around the
burrower, or internally through the digestive tract. The backfill may therefore contain both faecal
and nonfaecal components.

Boundary

The sharp or diffuse interface between the host sediment and the bioturbation structure. In contrast,
the outline or margin (herein used interchangeably) are more general terms describing the outermost
part of the trace fossil and may include the boundary, wall, mantle, some other peripheral structure,
or a combination of these structures.

Branching

Four types of branching (sensu lato) have been recognized (D’Alessandro and Bromley 1987). Of
these, ‘false’ and ‘secondary successive’ branching are applied to trace fossils that only apparently
ramify and are not truly branched.

Wall

A feature actively constructed by a burrowing organism to help provide the animal with temporary
or permanent protection from the external environment. In more permanent dwelling burrows the
wall might seal off permeability, allow canalization of the irrigation stream (Schafer 1956) or prevent
burrow collapse. In backfilled structures the need for a constructional wall is of little importance,
since the ‘burrow’ is not permanent and is immediately closed off behind the producer. Note that
the simple excavated boundary of a burrow is not a wall in the true sense because there has been
no active construction, only excavation. A true wall may be simple or complex, thin or thick, and
may be composed of compacted sediment, a sediment lining, pelletal, faecal, skeletal or vegetational
material, or a combination of such (Text-fig. 1a-b).

Lining

A type of wall structure formed by the active or passive attachment of typically fine-grained material
to mucus, applied by the producer to the interior side of its burrow. It is observed as the inner part
of a complex wall or, alternatively, as the only wall structure produced (Text-fig. 1a, and Bromley
1990, p. 20, fig. 2.7). A lining in a backfilled structure would probably only be actively produced as
a by-product of selective sediment sorting by the burrower (Clifton and Thompson 1978), or
passively accumulated on mucus secreted locally to assist passage of a soft-bodied organism through
the substrate.

Mantle

The outer zone of a two-zoned burrow fill (Text-fig. 1b). Heinberg (1974, p. 10) explained that this
feature is formed by a burrowing organism’s hydrostatic anchor for the purpose of forward
locomotion. Mantles are not constructed as insulation against external conditions or to ease passage
through the substrate, but are actual locomotory evidence of such passage. Therefore, as pointed

KEIGHLEY AND PICKERILL: BEACON ITES

307

text-fig. I. Morphological features of burrows, a, various types of wall construction (note that a lining is a
type of wall), b, various types of unwalled burrow margins; (i) and (ii) are backfilled structures and hence not
‘burrows’ in the strictest sense; (iii) is a simple excavated burrow; no construction has occurred at the burrow
boundary and so it is classed as being ‘unwalled’, c, styles of meniscate backfill.

out by Heinberg (1970) and reiterated by Bromley (1990, p. 149), the mantle is ‘conceptually distinct
from a true burrow wall'.

Meniscus

This term is utilized in a more restrictive sense than its Greek derivation of ‘shaped like a crescent
moon’, being limited to a transversely-oriented, arcuate to almost chevron-shaped, internal
interface that is observed on a surface trail in plan view, or in axial cross-section in a burrow or
backfilled structure. In this definition, use of the term ‘internal interface' emphasizes the two-
dimensional nature of a meniscus. It has no longitudinal dimension (i.e. it appears as a thin line
running transversely across the structure). The concave side marks the direction in which the
producer is travelling (Text-fig. lc). The meniscus is produced by the termination of an episode of
material backfilling behind the burrower.

Between successive menisci, the amount of backfilling that can occur is variable (Text-fig. lc). If
the amount of material processed is large, a bullet-shaped compartment, or meniscate packet , of
material is produced. With less material present between successive menisci, a three-dimensional,
dish-shaped compartment forms. In cross-sectional view, menisci converge at the margin of the
structure forming cusps, and the general appearance is that of a crescent (meniscate segment). A
third style of meniscate backfill comprises a non-compartmentalized meniscate backfill , where

308

PALAEONTOLOGY. VOLUME 37

menisci are too densely, or diffusely, stacked for separate segments or packets to be recognized. A
packet is most likely the result of a single excretory event (whether the packet contains visible
organic waste in the form of pelleted aggregates, or not), whereas non-compartmentalized backfill
may represent more continuous external backfilling (with or without scattered, intermixed, faecal
pellets). Potentially, the three types of meniscate fill are part of an intergradational sequence
dependent on how much material is processed, stuffed or excreted, and compacted behind the
animal as backfill at any one time between periods of forward locomotion. Meniscate backfill may
be homogeneous whereby the backfill (compartmentalized or not) is essentially of the same uniform
composition (sediment or faecal) throughout the structure. Alternatively, the backfill may be
heterogeneous , whereby more than one type/grain-size of internally or externally processed
sediment (and/or faecal material) is present or was presumed originally to be present before being
weathered out. Physical transport outside the body of the animal may itself provide for alternating
meniscate composition by way of the physical or compositional sorting of the sediment by the
organism (D'Alessandro and Bromley 1987; Pickerill 1989).

DIAGNOSTIC CRITERIA OF BEACONITES

The nature of the margin in a burrow or backfill structure, together with the presence or absence
of branching, are the significant primary diagnostic criteria at the ichnogeneric rank for the
distinction of most meniscate trace fossils, as emphasized by D’Alessandro and Bromley (1987),
D’Alessandro et al. (1987) and Bromley (1990). Factors such as lithology or geographical
distribution should not be considered as criteria for identifying Beaconites or any other ichnotaxon,
because they are contrary to the concept of ichnotaxobases (Pemberton and Frey 1982; Bromley
1990; Pickerill 1994). The following discussion concentrates on the true nature of the burrow outline
in Beaconites , and attempts to clarify problems in the diagnosis and differentiation of
morphologically similar ichnogenera, particularly Taenidium and Ancorichnus .

Beaconites as a walled meniscate trace fossil

Although based solely on photographs rather than specimens or field observations, Beaconites was
validly erected by Vialov (1962, p. 727). In the preamble to his diagnosis, he stated that the walls
were ‘fine’ and ‘clearly distinguishable’. Although Vialov’s definition of a ‘wall’ was not provided,
his comments are best interpreted as meaning an actively constructed wall comprised of fine-grained
material, or a tubular lining, as the terms are utilized herein. Bradshaw (1981, p. 630), in emending
the ichnogeneric diagnosis, similarly noted that the ichnotaxon was distinguishable as ‘tubular
burrows’ with a smooth burrow lining. This same construction, occurring as a sand lining, was also
noted in her emended diagnosis of the type ichnospecies B. antarcticus Vialov, 1962. Such
characteristics have also been used consistently by other workers (e.g. Webby 1968; O’Sullivan et ah
1986; Woolfe 1990; Sarkar and Chaudhuri 1992) in their interpretation of Beaconites. The
presence of a lined wall is therefore a primary diagnostic criterion for the definition of this trace
fossil.

Despite Vialov’s (1962) original diagnosis and Bradshaw’s ( 1981 ) emendment, several subsequent
authors (e.g. Frey et ah 1984; Squires and Advocate 1984) considered the ichnogenus to be unlined
or unwalled. Accordingly, they suggested that Beaconites might be a junior synonym of some other
unwalled meniscate burrow. D’Alessandro and Bromley (1987, p. 751) also initially considered the
ichnotaxon as dubious, ‘...having a weak wall or none at all, and it should probably be included
in Taenidium' . In contrast, they later commented that B. antarcticus seemed to have a wall, and that
details of the burrow boundary, from photographs, topotypic material, and descriptions, were
unclear (D’Alessandro and Bromley 1987, p. 757). Examination of unweathered material was
therefore considered to be necessary before the relationship between Beaconites and Ancorichnus or
Taenidium could be clarified. It is, however, difficult to visualize how examination of unweathered
specimens would alleviate nomenclatural difficulties since, if Bradshaw’s (1981) specimens are

KEIGHLEY AND PICKERILL: BEACON ITES

309

considered to be too badly weathered, and Vialov’s (1962) material was never collected, it cannot
be ascertained what type of marginal structure constituted the original Beaconites (and assignment
of new material would depend on the diagnoses of the original ichnotaxa). As with many ichnotaxa
proposed in the nineteenth century, ichnologists have only the original diagnosis and illustrations
to work with. From both the validly introduced original and emended diagnoses, Beaconites must
remain a walled meniscate burrow.

The confusion over whether Beaconites was a walled or unwalled burrow probably persisted not
only due to poorly defined terminology but also because of inclusion within it of large meniscate
forms. Gevers et al. (1971) were the first to consider large meniscate burrows as B. antarctieus ,
possibly on the mistaken belief that Vialov’s (1962) specimens could attain a diameter of 150 mm.
In comparing their giant forms with other large burrows, they noted (Gevers et al. 1971, p. 83) that
their largest examples ‘compare in dimensions with the largest (150 mm) of B. antarctieus as
described by Vialov’. The largest specimens described by Vialov (1962) were, however, no larger
than 15 mm in diameter. Gevers et al. (1971) also suggested that nomenclatural distinction from
Vialov’s (1962) specimens might be justified at the ichnospecific level, but that this distinction
should be based upon significant differences in both the thickness of the meniscate packets, and in
the ‘septal’ (meniscate) shape. One of the differences they failed to recognize, however, was the
contrast in the outline of the trace fossil between their material and the type specimen of
B. antarctieus. They interpreted the coalescing ‘ridges’ (individual meniscate segments) at the
margin as a lining, or ‘welt’, that they considered to be the erosional remnants of ‘outer
consolidated tubes’. The ‘welts’ were noted to thicken where the burrow curved sharply and the
‘septal’ ridges were close together. Photographs of these large forms, which they noted as
containing a marginal furrow, similarly illustrate the burrow outline being formed by the merging
of transverse meniscate ridges that were produced by backfill (Gevers et al. 1971, pi. 18, figs 1, 2,
4). The large forms apparently had no true lining or other type of wall structure. In comparison,
their small burrows had a distinct lining and hence were walled (Gevers et al. 1971, pi. 18, fig. 3)
and correctly described as B. antarctieus.

Recognizing that two distinct morphologies were being included within the same ichnospecies,
Bradshaw (1981) followed the suggestion of Gevers et al. (1971) and introduced a second
ichnospecies of Beaconites. The description (= diagnosis) of B. barretti Bradshaw, 1981 (p. 631)
stated that the sediment compartments ‘...may merge laterally to form a crude burrow lining...'
when the meniscate segments ‘...meet the burrow wall at an acute angle’. However, in her
specimens, the ‘lining’ of B. barretti is produced by the backfill, specifically the peripheral margins
of the meniscate segments. The producer simply excavated and backfilled its burrow. No wall was
constructed. Neither is the presence of a true wall lining, or any other sort of wall structure,
supported by her illustrations (Bradshaw 1981, figs 17 and 18). B. barretti , both as defined and
illustrated, is not walled. Indeed, of all the large meniscate trace fossils previously or subsequently
described and figured as B. antarctieus , B. barretti , or simply Beaconites (e.g. Gevers et al. 1971 ;
Hantzschel 1975; Pollard 1976; Allen and Williams 1 98 1 z?, 19816; Bradshaw 1981; Graham and
Pollard 1982; Narbonne 1984; Eagar et al. 1985; Briick 1987; Dam and Andreasen 1990; Woolfe
1990; Pearson 1992; Tegan and Curran 1992), none appears either lined or walled. This clearly
contrasts with both the original and emended diagnosis of the ichnogenus. These large backfilled
burrows cannot be included within Beaconites and must, instead, be incorporated within an
alternative meniscate ichnotaxon, or re-introduced as a new ichnogenus. The former option is
adopted for the reasons discussed below.

Unwalled meniscate trace fossils : large ‘ Beaconites ’ forms and Taenidium

To determine to which ichnotaxon the large ‘ Beaconites ' forms should be assigned, it is necessary
to review briefly the current status of simple, unwalled, backfilled burrows. Debate has been going
on since the mid-nineteenth century as to the potential synonymies of the meniscate trace fossils
Muensteria , Keckia Glocker, 1841, Taenidium , and Beaconites (e.g. Fischer-Ooster 1 858 ; Heer 1 877 ;

310

PALAEONTOLOGY, VOLUME 37

Ancorichnus

ox

A. ancorichnus

Taenidium

H3D

T. serpentinum

T. cameronensis

text-fig. 2. Currently accepted ichnospecies of Ancorichnus , Beaconites , and Taenidium , showing the variation
that might occur in backfill morphology for each ichnospecies.

Schroter 1894; Liburnau 1900; Wilckens 1947; Frey et al. 1984; Squires and Advocate 1984;
O'Sullivan et al. 1986; McCann and Pickerill 1988). Of most importance, D' Alessandro and
Bromley (1987, p. 747) considered Muensteria a name unavailable for trace fossils, concluding that
"...on the basis of its first ichnospecies Taenidium is available for the unbranched ichnospecies of
Muensteria...' Branched, annulate burrows of Taenidium were transferred to Cladichnus
D’ Alessandro and Bromley, 1987.

Several of the actions of D’Alessandro and Bromley (1987) are, however, in need of clarification.
As Sternberg’s (1833) original Muensteria included algae, coprolites and possible specimens of
Chondrites Sternberg, 1833, it was therefore erected as a heterogeneous ichnogenus, and thus
invalidly introduced. It would, however, remain as an available name (International Code of
Zoological Nomenclature 1985, Articles 1 0—1 4, and 17). Although the heterogeneous nature of
Muensteria was addressed in the emendments of Fischer-Ooster (1858), many of his actions,
particularly with regards to the introduction of three ‘subgenera', were again probably invalid as
dictated by today’s I.C.Z.N. D’Alessandro and Bromley (1987) considered Taenidium as an
unbranched ichnotaxon because the type ichnospecies was diagnosed as unbranched. This is not the
case because Heer (1877) did not consider branching as an important diagnostic criterion. The
original diagnosis of the ichnogenus (as opposed to that of the type ichnospecies) clearly stated that
the taxon was ‘ rarius ramosa' (rarely branching) in the sense that one or two ichnospecies, such as
T. fischeri Heer, 1877, could branch. Additionally, according to Heer (1877), Wilckens (1947), and
McCann and Pickerill (1988), amongst others, Taenidium was differentiated from forms ascribed to
Muensteria because the latter may lack packeting and a ringed, or annulate, boundary and instead,
contain only simple, non-compartmentalized meniscate (‘runzelig quergestreift ’ (transversely

KEIGHLEY AND PICKERILL: BEACON ITES

31 1

TRACE FOSSIL

AUTHORS

Recorded burrow widths (mm)

) 50 100 150 200 250

....I 1 1 1 1

'Taenidium'

Smith et at. 1993

n (8-15 mm) .

'Taenidium

D’ Alessandro et at. 1993

(12-25 mm).

T. serpentinum

Chamberlain 1977

B (1-2 mm).

T. serpentinum

Dam 1990a

B (5-10 mm).

T. salanassi

D’ Alessandro and Bromley 1987

IB (4-14 mm).

T. cameronensis

Brady 1947

„ (12-18 mm).

T. barred i

Gevers et at. 1971

T. bar r ell i

Ridgway 1974

T. barretti

Allen and Williams 1981

(30-250 mm).

T. barretti

Bradshaw 1981

T. barretti

Graham and Pollard 1982

T. barretti

Briick 1987

(150-230 mm).

T. barretti

This study

text-fig. 3. Recorded widths of Taenidium and its ichnospecies from selected literature. As these are not
necessarily diameters, the data may contain some positive skew (Graham and Pollard 1982). In addition to the
75 mm wide specimen from the Port Hood Formation (Plate 1, figs 5-6), this study also measured a total of
56 specimens from the Perry Formation (see also Plate 1, figs 2^4). Over a discontinuous area of the same
bed (approximately 6 m by 2 m), burrows range in width from 5 to 51 mm (mean = 18-1 mm, standard

deviation = 9-7).

striped); Wilckens 1947) backfill. The emended diagnosis of Taenidium (D’ Alessandro and Bromley
1987) similarly stated that these burrows should contain a segmented fill articulated by meniscus-
shaped partings (i.e. the backfill is distinctly compartmentalized, consistent with the diagnoses of
Heer 1877 and Wilckens 1947). Contrary to this, from D'Alessandro and Bromley's (1987, p. 747)
sweeping statement quoted above, and their discussion of other potentially synonymous ichnotaxa,
it would appear that all unbranched forms of Muensteria , and other meniscate ichnotaxa, are to be
included within Taenidium , regardless of the style of backfill. This might be seen as an excessive
lumping of ichnotaxa but, as discussed earlier, segments and simple backfill in particular may
potentially intergrade within a single burrow. Differentiation of separate ichnogenera for variations
in the style of backfill may therefore be considered inappropriate.

Despite the problems outlined above, to promote ichnotaxonomic stability we follow
D’Alessandro and Bromley (1987) and include all radially branched, and potentially palmate,
ichnospecies of both Muensteria and Taenidium within Cladichnus. Similarly we follow their
suggestion to include unbranched ichnospecies of Muensteria within Taenidium (note that different
specimens may cross each other giving a falsely branching appearance, Plate 1, fig. 1). Variation in
the type of backfill contained within these, and other potentially synonymous, unbranched,
meniscate burrows is therefore relegated to an ichnospecies-level taxobase.

Another emendment to the diagnosis of Taenidium proposed by D'Alessandro and Bromley
(1987) is, however, still required. These authors potentially allowed for both unlined and thinly lined
burrows to be included within the ichnogenus. From their diagnosis, the production of a thin,
continuous lining can be considered a morphological feature comparable to no lining being present,
and not worthy of ichnogeneric distinction. The production of a thick lining, in contrast, is inferred
to be a morphological feature different from that of a thin lining and significant enough to warrant
a separate ichnogeneric name. What is construed as ‘thin’, or for that matter ‘thick’, is arbitrary,
and such a division cannot improve ichnotaxonomic stability. The presence of any form of
continuous burrow lining, thick or thin, should be considered a morphologically similar criterion,
and different to that of no lining. The presence or absence of a continuously lined wall is therefore
an important ichnotaxobase. Consequently, as a further emendment to the diagnosis of Taenidium ,

312

PALAEONTOLOGY, VOLUME 37

we propose that this ichnogenus should be restricted to unwalled, and thus unlined, burrows. Such
an emendment would also reinforce the differences between Beaconites and Taenidium , as discussed
in the previous section, and avoid further lumping of ichnotaxa.

Accordingly, although D’ Alessandro and Bromley (1987) concluded that Beaconites should
probably be included within Taenidium , only B. barretti is regarded as synonymous. Without
exception, B. barretti burrows are unbranched, meniscate, cylindrical and, above all, unlined. Such
burrows are distinguished as a distinctive ichnospecies, specifically T. barretti (Bradshaw), on the
basis of the heterogeneous, thinly segmented or non-compartmentalized, arcuate meniscate fill. The
style of meniscate fill in these burrows is clearly at variance with the three ichnospecies of Taenidium
considered by D’Alessandro and Bromley (1987), all of which are distinctly packeted (Text-fig. 2)
but this is a legacy of the latter authors' relegation of style of meniscate fill to an ichnospecies-level
diagnostic criterion.

The size of a burrow has, in the past, been used as a significant criterion in formulating new
ichnogenera (e.g. Megagyrolithes Gaillard, 1980) as well as a secondary criterion in distinguishing
ichnospecies within an ichnogenus (e.g. Helminthopsis Heer, 1877). The range of widths measured
for specimens of T. barretti , although typically much greater, overlap to some degree those widths
previously measured in other ichnospecies of Taenidium and Muensteria (Text-fig. 3). Therefore,
despite the measured widths of T. barretti probably including some positive skew, resulting from the
measurement of widths on bedding planes as opposed to true burrow diameters (Graham and
Pollard 1982), it would be unwise for these burrows to be considered distinct at the ichnogeneric
level based on size. T. barretti observed at various localities in eastern Canada further illustrate
the variability in size that can be encountered within otherwise morphologically similar specimens
(Plate 1).

Vertical escape and vertical adjustment structures

Whether typically large, vertically oriented trace fossils containing distinctive arcuate, concave
upwards, nested menisci should be attributed to a previously established meniscate backfilled
burrow, such as Beaconites , was questioned by Bridge et cd. (1986). Most commonly, such trace
fossils have been left in open nomenclature and described simply as ‘escape structures’ or
‘equilibrium structures’ (e.g. Eagar et cd. 1985; Sarkar and Chaudhuri 1992). Occasionally vertical
‘burrows’ have been erroneously named as Beaconites'. the structures of Bridge and Gordon (1985),
Bridge and Droser (1985), and Berg (1977) do not contain a wall, whereas those of Allen and
Williams (1981a, 19816) and Briick et al. (1985) additionally do not have an observable meniscate
structure.

The morphological characteristics of a trace fossil provide the only feasible diagnostic criteria for
distinguishing ichnotaxa. The behavioural interpretation of a trace fossil is not a valid criterion,
though morphological features may be indicative of one particular behaviour, or a variety of
behaviours: meniscate backfill containing faecal matter or meniscate packeting is typical of

EXPLANATION OF PLATE 1

Figs I -6. Taenidium barretti (Bradshaw). Burrows of variable size and variable distinctiveness of meniscate fill ;
1-4, preserved in convex epirelief; 5-7, preserved in convex hyporelief; they are ubiquitously non-
compartmentalized and meniscate and much less distinct where burrow-fill is of uniform composition.
1, Alma, southeast New Brunswick, Canada; alluvial fan deposits, Hopewell Conglomerate, Lower
Carboniferous; x 0-6. 2-4, McCann Cove, southwest New Brunswick, Canada; partly calichified alluvial fan
deposits. Perry Formation. Devonian; 2, x0 4; 3, x05; 4, xO-9. 5, near Port Hood, Cape Breton Island,
Nova Scotia, Canada; fluvial bar sandstones, upper Port Hood Formation, Upper Carboniferous; xO-2. 6,
oblique view of same specimen as in fig. 5.

All field photographs, specimens not collectable.

PLATE I

KEIGHLEY and PICKERILL, Taenidium

314

PALAEONTOLOGY, VOLUME 37

fodinichnia (Bromley 1990), but backfilled burrows lacking distinct faecal content or meniscate
packeting may represent repichnia, equilibrichnia of Bromley (1990), or even fugichnia. Vertically-
upward oriented, non-compartmentalized, meniscate structures may be repichnia, while oblique,
lateral or vertically downward-oriented structures may represent adjustment or escape from lower
water tables, desiccation, erosion or predation (cf. Sarkar and Chaudhuri 1992, fig. 5; if way-up was
not known, how would the two structures be differentiated?). Since Taenidium has never been
restricted to a particular orientation with respect to stratification, simple, vertically-upward
oriented, backfilled structures with arcuate menisci can be accommodated within this ichnogenus.
Vertically oriented 'escape" structures that contain irregularly patterned, downward-deflected
backfill (e.g. Pienkowski 1985, plate 1b; Bromley 1990, fig. 5.8c; Gierlowski-Kordech 1991, fig. 7)
are distinguishable from Taendium and must, for the present, be retained in open nomenclature.

Beaconites and other waded meniscate trace fossils

Since Beaconites has a distinct wall structure, the relationship with other supposedly walled
meniscate trace fossils, particularly Scoyenia and Ancorichnus , must also be considered. Scovenia ,
originally described in botanical terms, has subsequently been considered a walled, meniscate
burrow that possesses longitudinally striated linings (Frey et cd. 1984). In this sense Scoyenia and
Beaconites are potential taphonomic variants of the same burrow, taxonomic classification
depending on the presence or absence of such longitudinal striations. Bromley (pers. comm. 1993)
suggested that Scoyenia is actually unlined and unwalled, with striations produced on the simple
excavated margin of a transient burrow. If this is the case, it would mean that Scoyenia and
Taenidium are potential taphonomic variants. The separate identity of Scoyenia is, however,
retained for ichnotaxonomic stability.

When originally proposed, Ancorichnus was monospecific. The type, A. ancorichnus Heinberg,
1974, was described as an undulating 'cylindrical meniscus filled tunnel with a distinct mantle". The
mantle was considered as the outer part of a two-zoned burrow fill and had, as an internal structure
or ornament, a distinct orientation of mica grains. The mantle, with its internal ornamentation, was
not believed to have been formed by the construction of a wall but rather by the burrowing
organism’s hydrostatic anchor for the purpose of forward locomotion (Heinberg 1974, p. 10). As
noted by Bromley (1990, p. 149), this type of margin is 'conceptually distinct from a true burrow
wall".

Heinberg (1974, p. 9) also recognized that, in some cases, the mantle could be weathered out and,
instead, be represented as a groove on each side of the meniscate core-fill. In such cases, the mantle
might be mistaken for a weathered-out wall. Since no internal ornament would be preserved in the
grooves, there would be essentially no difference between Ancorichnus and those burrows described
by Vialov (1962) or Bradshaw (1981) as Beaconites. Depending upon taphonomic variability,
problems may therefore also exist in the distinction of Ancorichnus and Beaconites. Nevertheless, the
presence or absence of an oblique or transverse internal fabric in the marginal structure of a
meniscate trace fossil will best determine whether it has a mantle or a lining and whether it should
therefore be assigned to Ancorichnus or Beaconites.

Further ichnotaxonomic problems have arisen because of the proposed changes to Ancorichnus
by Frey et cd. ( 1984, p. 514). Linings, mantles and walls became less rigidly defined in their emended
diagnosis: ‘...relatively thick, unornamented wall linings surrounding well-developed, meniscate
burrow fills. Distal ends of menisci may blend directly into the wall structure." Essentially, the
mantle in A. ancorichnus was considered as a wall or a wall lining, and its internal structure was
relegated to a feature of ichnospecific significance. Accordingly, A. capronus Howard and Frey,
1984, and A. coronus Frey et cd.. 1984, two simply lined burrows, were also accommodated within
the ichnogenus, even though they did not have a distinctly structured mantle.

Although Frey et cd. (1984) made careful comparisons with several other meniscate burrows with
which Ancorichnus and Scoyenia could be confused, synonymy with Beaconites was dismissed
because they considered Beaconites an unlined burrow (Frey et cd. 1984, p. 517). However, as

KEIGHLEY AND PICKERILL: BEACONITES

315

Beaconites is lined, acceptance of the emended diagnosis of Ancorichnus by Frey et al. (1984) would
mean that there is no primary criterion to distinguish it from Beaconites , and it would therefore
become a junior synonym. By rejecting the emendment of Frey et al. (1984) in favour of the more
precise description originally provided by Heinberg (1974), the important distinguishing
characteristic of Ancorichnus is retained, namely the presence of a distinctly structured mantle. It
is morphologically and behaviourally distinct, because the burrow outline is produced by a method
akin to active backfill; in Beaconites the outline is a constructional lining. Ancorichnus remains
monospecific, containing only A. ancorichnus. D'Alessandro et al. (1993) suggested that A. coronas
may not have a discrete wall structure, but rather that its marginal structure was formed by partial
overlapping of contiguous menisci (as is the case with B. barretti). They suggested placing the
ichnospecies within Taenidium. If this were done, A. coronas would be a junior synonym of
T. barretti , as revised herein. However, the original diagnosis of the ichnospecies is followed by Frey
et al. (1984), and A. coronas is considered to be walled. Even so, it could then be considered a junior
synonym of B. antarcticus because of its distinct, unstructured wall lining. This contribution
proposes that the ichnospecific identity of A. coronas be retained, albeit within the walled
ichnogenus Beaconites. As with Taenidium , variations in fabrication of meniscate fill can be used to
differentiate ichnospecies. Meniscate packets of a thickness approximately equivalent to the
diameter of the burrow typify B. antarcticus , whereas most specimens of what have previously been
called A. coronas have regularly arranged, heterogeneous, longitudinally short meniscate packets or
even segments (Text-fig. 2). The latter forms are therefore identified separately as B. coronas (Frey
et al. 1984). A. capronus , by way of its distinctive chevron pattern of meniscate infill within a thin,
unstructured wall lining, is similarly retained as a distinct iclmofossil, becoming B. capronus
(Howard and Frey, 1984).

SYSTEMATIC PALAEONTOLOGY

As this contribution deals primarily with the clarification of Beaconites , and its relationship to the
morphologically similar Ancorichnus and Taenidium , included synonymy is restricted to these three
ichnotaxa. Only a brief systematic treatment of Taenidium was provided by D’Alessandro and
Bromley (1987), and we have now expanded and updated their listing. In numerous cases, ichnotaxa
have only been illustrated schematically (in varying degrees of detail) and with limited accompanying
description. Such references can only be considered as uncertain synonymies (and are prefixed '?')
and typically are only identified to the ichnogeneric level. Additionally, since Taenidium is now
considered to be the first valid name for specimens of Muensteria , former ichnospecies of the latter
that are now of indeterminate ichnospecific designation within Taenidium , or are reassigned to other
ichnotaxa, are included within the ichnogeneric synonymy of Taenidium.

non 1984
non 1984
non 1984

non 1985

non 1985

non 1985

11986
non 1987
non 1989
non 1990

Ichnogenus ancorichnus Heinberg, 1974

Ancorichnus capronus Howard and Frey, p. 201, figs 2-3 [= Beaconites capronus ].

Ancorichnus capronus Howard and Frey; Frey et al., p. 514 [= Beaconites capronus ].
Ancorichnus coronus Frey et al., p. 511, figs Id [copy of Stanley and Fagerstrom 1974, fig. 6], 1e,
3a-c [= Beaconites coronas],

Ancorichnus capronus Howard and Frey; Frey and Howard, p. 373, figs 5.6, 5.8, 16.3d
[= Beaconites capronus],

Ancorichnus capronus Howard and Frey; Frey and Howard, p. 122, fig. 2 [copy of Howard and
Frey 1984, fig. 2; = Beaconites capronus].

Ancorichnus capronus Howard and Frey; Frey and Pemberton, p. 90, fig. 26 [copy of Howard
and Frey 1984, fig. 2; = Beaconites capronus],

Ancorichnus isp. ; Valenzuela et al., p. 129.

cf. Ancorichnus coronus Frey et al. ; D’Alessandro et al., p. 285, fig. 2 [? = Taenidium barretti],
Ancorichnus capronus Howard and Frey; Martino, p. 393, fig. 5.5 [= Beaconites capronus].
Ancorichnus capronus Howard and Frey; Frey, p. 204, fig. Id [= Beaconites capronus].

316

PALAEONTOLOGY, VOLUME 37

LOCATIONS *authors

AGE

ENVIRONMENT

Alluvial Fan

Lacustrine

Floodplain

/Overbank

Crevasse Splay
/Levee

Fluvial

Distributary

Channel

Mouth Bar
/Intertidal

Inner Shelf
(a.s.w.b.)

Outer Shelf
(b.s.w.b.)

Submarine fan
/Abyssal

Nebraska, USA ^

S England

4-13

Utah, USA

14

Alberta, Canada

15-16

Jameson Land, Greenland

17,18

Milne Land, Greenland
19

West Virginia, USA

20-24

S Victoria Land Antarctica

Miocene

Oligocene

Eocene

Cretaceous

Jurassic

Jurassic

Carboniferous

Devonian

■H-H-f

KEY: A. ancorichnus B. antarcticus B . capronus ■■■■■■ B coronas

text-fig. 4. Temporal and environmental distribution of ichnospecies of Ancoriclmus and Beaconites. Only
references to confidently assigned ichnospecies and of known environment and age are included. Authors: 1,
Stanley and Fagerstronr 1974; 2, Frey et al. 1984; 3, Daley 1968; 4, Howard 1966; 5, Frey and Howard 1970;
6, Howard 1971 ; 7, Howard 1972; 8, Frey and Howard 1982; 9, Howard and Frey 1984; 10, Frey and Howard
1985a; 1 1, Frey and Howard 19856; 12, Frey 1990; 13, Frey and Howard 1990; 14, Pemberton and Frey 1984;
15, Heinberg 1970; 16, Heinberg and Birkelund 1984; 17, Heinberg 1974; 18, Fiirsich and Hemberg 1983; 19,
Martino 1989; 20, Vialov 1962; 21, Haskell et al. 1965; 22, Webby 1968; 23, Gevers et al. 1971 ; 24, Bradshaw

1981.

non 1990 Ancoriclmus capronus Howard and Frey; Frey and Howard, p. 808, fig. 12 [copy of Frey 1990,
fig. 7d], fig. 13.1 [= Beaconites capronus ].

non 1991 Ancorichnus coronus Frey et al.', Acenolaza and Buatois, p. 96, pi. 2.1 [= Taenidium barretti ],
pi. 2.3 [indeterminate],

non 1993 Ancorichnus coronus Frey et al. : Acenolaza and Buatois, p. 188, fig. 4d [copy of Acenolaza and
Buatois 1991, pi. 2.1; = Taenidium barretti ].

non 1993 Ancorichnus aff. coronus Frey et al.: Mikulas, p. 106, fig. 3 d [= indeterminate].

Diagnosis. Cylindrical, weakly sinuous, sub- to horizontal burrow containing a central meniscate
fill and a structured mantle (after Heinberg 1974).

Type ichnospecies. Ancorichnus ancorichnus Heinberg, 1974 (by monotypy)

Remarks. According to Heinberg and Birkelund (1984) Muensteria (sensu lato) includes the
ichnogenus Ancorichnus. Whether they regarded Ancorichnus as an ichnosubgenus of Muensteria ,
or a synonym of Muensteria , is uncertain. However, Muensteria is now included within the unwalled
ichnogenus Taenidium and, while both Ancorichnus and Taenidium are unwalled, the latter
comprises a simple, one-stage backfill, whereas Ancorichnus has a two-stage fill comprising an inner
meniscate fill and an outer mantle.

Ancorichnus is considered to be monospecific. A. capronus Howard and Frey, 1984, and
A. coronus Frey el al. 1984, are now included within Beaconites.

Ancorichnus ancoriclmus Heinberg, 1974

1970 Meniscus tunnel; Heinberg, p. 230, fig. 3c.

*1974 Ancorichnus ancorichnus Heinberg. p. 7, figs 1a, 2^4, 9a.

KEIGHLEY AND PICKERILL: BEACONITES

317

1983 Ancorichnus ancorichnus Heinberg; Fursich and Heinberg, p. 94, fig. 7. II.

1984 Muensteria Sternberg; Heinberg and Birkelund, p. 365, fig. 10b.

71990 Ancorichnus ancorichnus Heinberg; Dam, p. 121, figs 4, 7a.

11990 Ancorichnus ancorichnus Heinberg; Dam, p. 221.

1990 Ancorichnus ancorichnus Heinberg; Bromley, p. 149, fig. 9.3.

Diagnosis. As for ichnogenus.

Remarks. The mantle in Ancorichnus is an integral part of the burrow fill, produced by the
hydrostatic anchoring of the burrow producer. The mantle is not a wall structure. To avoid possible
confusion with Beaconites, a structured mantle should be recognized or confidently inferred ( contra
the emendment by Frey et al. 1984).

A. ancorichnus has only been recorded from the Jurassic of Greenland (Text-fig. 4). This is
probably because excellent preservation of the fabric in the mantle is required for it to be
recognizable, and weathering of rock surfaces may be retarded in the high arctic. The mantle
typically exhibits a structure formed by the preferred orientation of the sediment, and usually this
is defined by oriented mica flakes (Heinberg 1974, fig. 4; Bromley 1990, fig. 9.1). Likewise, in
Heinberg’s (1974) specimens, the weakly arcuate menisci of the inner fill are defined by oriented
mica flakes.

Determining whether a marginal structure is a mantle or a wall is difficult. However, the fabric
in the mantle is oriented at an angle to the burrow boundary, as opposed to being normal to the
direction of the burrow, which would be more characteristic of an applied wall lining. If internal
structure in the mantle cannot be recognized, assignment to Ancorichnus is of course, problematic.
Taphonomic variants (e.g. Dam 1990<r/, 1990/d may not contain a visibly structured mantle, but
Dam's material was differentiated from Beaconites because the latter typically contains a lining
(usually fine-grained) that differs in composition from the surrounding substrate. Dam (1990<r/,
fig. 4) diagrammatically illustrated A. ancorichnus with Jamesonichnites heinhergi Dam, 1990, as a
compound specimen ( sensu Pickerill 1994) and suggested that the two types of burrow represent
different patterns of behaviour by the same organism.

non 1975

? 1976
non 1981
non 1984
non 1985
non 1985

? 1985
non 1987

? 1988
non 1988
non 1990
non 1990
non 1992
non 1992

Ichnogenus beaconites Vialov, 1962

Beaconites Vialov; Hantzschel, p. W45, fig. 28.1 [copy of Gevers et al. 1971, fig. 2 = Taenidium
barretti],

Beaconites Vialov; Scott et al.. p. 4.

Beaconites barretti Bradshaw, p. 630, figs 17-18 [= Taenidium barretti ].

Beaconites Vialov; Narbonne, p. 409, fig. 8g [= Taenidium cf. serpentinum).

Beaconites barretti Bradshaw; Bridge and Gordon, p. 173, fig. 8a-b [vertical burrow].
Beaconites barretti Bradshaw ; Bridge and Droser, p. 1 54, fig. 8b [indeterminate vertical and
horizontal structures],

Beaconites barretti Bradshaw; Bridge and Nickelsen, p. 187.

Beaconites barretti Bradshaw; Bruck, p. 259, figs 4-5 [= Taenidium barretti ].

Beaconites Vialov; Pollard, p. 339.

Beaconites barretti Bradshaw; Bradshaw and Webers, p. 787 [= Taenidium barretti],
Beaconites isp. ; Dam and Andreasen, p. 208, fig. 1 1 b [= Taenidium barretti],

Beaconites barretti Bradshaw; Woolfe, p. 302, fig. 3 [ = Taenidium barretti],

Beaconites barretti Bradshaw; Tegan and Curran, p. A 154 [= Taenidium barretti],

Weaconites Vialov; Pearson, p. 129, figs 3d-e [3d = Taenidium barretti, 3e indeterminate].

Emended diagnosis. Small, cylindrical, unbranched, walled, meniscate burrow. Straight or sinuous,
horizontal or more rarely inclined or vertical. Weakly to strongly arcuate meniscate packets or
segments enclosed by distinct, smooth and unornamented burrow linings.

Type iclmospecies. Beaconites antarcticus Vialov, 1962 (by monotypy).

318

PALAEONTOLOGY, VOLUME 37

Remarks. Beaconites has distinct but structureless walls and, to date, has not been recognized with
a diameter exceeding 30 mm. Large, unwalled, unlined meniscate trace fossils should not be
included within this ichnogenus; B. barretti, is transferred to Taenidium. A. coronas and A. capronus
have been transferred to Beaconites as distinctive ichnospecies, since the structure surrounding their
central fill is a distinct wall or lining.

In addition to the records of Beaconites in the synonymy lists, other references to Beaconites have
been included in tabulations by Pollard et al. (1982), Pollard ( 1988), and Maples and Archer (1989).
Pollard et al. (1982) assigned the ichnofossils of Trewin (1976) and Berg (1977), amongst others, to
Beaconites. Berg’s (1977) specimens are indeterminate vertical burrows, whereas Trewin (1976)
figured and described only Scolicia Quatrefages, 1849, and Cruziana Orbigny, 1842. Pollard (1988)
reported the occurrence of Beaconites in mid-Jurassic cores from offshore Britain and also
attributed trace fossils described by Stanistreet et al. (1980), to this ichnogenus. These latter traces,
originally reported as Rhizo cor allium Zenker, 1836, do not have marginal tubes, thereby excluding
them from a rhizocoralliid designation. Assignment to Beaconites is also doubtful following the
comment of Stanistreet et al. (1980) that although the burrows are more than 200 mm wide, they
are only 10-20 mm in depth, suggesting that these traces are surficial trails, not cylindrical backfilled
structures. Maples and Archer (1989) included reference to Gevers and Twomey (1982) — a report
that mentioned structures of Haskell et al. (1965) — and reference to an abstract by Plume (1983),
that does not name any ichnofossils.

O’Sullivan et al. (1986) noted 'Beaconites- like’ burrows (that they distinguished from the
ichnogenus by a passive burrow fill) and Bamford et al. (1986) noted ' Beaconites- type’ burrows that
in fact contained distinct walls and lacked internal menisci. Allen and Williams ( 1 98 1 «) also
reinterpreted ‘mechanical structures’ of Dixon (1921), Allen (1961, 1963), Dineley (1966) and
Horne and Gardiner (1973), as large burrows attributable to Beaconites. Horne and Gardiner
(1973), however, correctly considered their specimens to be most similar to Kulindriehnus Hallam,
1960, a trace fossil subsequently placed in synonymy with Bergaueria langi (Hallam, I960) by
Pemberton et al. (1988). The examples of the other authors are of uncertain meniscate backfill.
Finally, other authors have included reference to Beaconites when commenting upon possible
synonymies for meniscate trace fossils. Frey et al. (1984) and Squires and Advocate (1984)
considered including it within other unlined meniscate burrows, and D’Alessandro and Bromley
(1987) within an ichnotaxon comprizing weakly walled or unwalled meniscate burrows.

*1962
1965
1968
1971
non 1971

71975
non 1976

11977

11977

11978

11978

11978

11979
non 1980

non 1981
non 1981

1981

Beaconites antarcticus Vialov, 1962

Beaconites antarcticus Vialov, p. 726, figs 9-10.

Beaconites antarcticus Vialov; Haskell et al.. p. 243.

Beaconites antarcticus Vialov; Webby, p. 1004. fig. 7.

Beaconites antarcticus Vialov; Gevers et al.. p. 81, pi. 18, fig. 3.

Beaconites antarcticus Vialov; Gevers et al.. p. 81, pi. 18, figs 1 2, 4 [= Taenidium barretti],
Meniscate burrow; Edwards, p. 7, fig. 1.

Beaconites cf. antarcticus Vialov; Pollard, p. 105 [= Taenidium barretti}.

Beaconites antarcticus Vialov; McKelvey et al., p. 829.

Beaconites antarcticus Vialov; Forsyth and Chisholm, p. 19.

Beaconites antarcticus Vialov; Allen, p. 88.

Beaconites antarcticus Vialov; Williams, p. 96.

Beaconites antarcticus Vialov; Allen and Williams, p. 121.

Beaconites antarcticus Vialov; Allen, p. 70.

Beaconites antarcticus Vialov; Bridge et al., p. 154, pi. 2d-e [2d = Bergaueria', 2i is
indeterminate].

Beaconites antarcticus Vialov; Allen and Williams, p. 23, fig. 6a [indeterminate vertical burrow],
Beaconites antarcticus Vialov; Allen and Williams, p. 255, figs 2, 4-1 1 [2, 4-5 are vertical
burrows; 6-11 = Taenidium barretti ].

Beaconites antarcticus Vialov; Bradshaw, p. 630, fig. 15 [7 fig. 16].

KEIGHLEY AND PICKERILL: BEACONITES

319

non 1982

21985
11985
non 1985
non 1985
non 1988
21990
11993

Beaconites antarcticus Vialov; Graham and Pollard, p. 259, figs 3a-b, 4a-c, 5a-b. [= Taenidium
barret ti\.

Beaconites antarcticus Vialov; Bridge and Gordon, p. 170, fig. 8c.

Beaconites antarcticus Vialov; Bridge and Nickelsen, p. 187.

cf. Beaconites antarcticus Vialov; Eagar et al ., p. 134, pi. 14a [= Taenidium barret ti].
Beaconites antarcticus Vialov; Brack et al., p. 87, figs 4a-b, 5a-b [vertical burrows],
Beaconites antarcticus Vialov; Gordon, p. 143, figs 4a-b [= Taenidium isp ]

Beaconites antarcticus Vialov; Woolfe, p. 302, figs 3-4.

Beaconites antarcticus Vialov; Ekes, p. 469.

Emended diagnosis. Small, cylindrical, unbranched, lined burrows, straight to slightly sinuous.
Burrow infill meniscate, typically heterogeneous packets of unequal thickness. Larger packets
slightly thicker to slightly thinner than overall burrow width, meniscate interfaces weakly to
moderately arcuate. Burrow lining is uniform, smooth and unornamented, commonly thick and
very distinct.

Remarks. As noted by D'Alessandro and Bromley (1987), B. antarcticus has sediment packets of
unequal length. Commonly, specimens are encountered where thicker packets have been
preferentially weathered out, preserving only thin compartments, or ‘septa' (e.g. Vialov 1962;
Gevers et al. 1971). The burrows of McKelvey et al. (1977), Forsyth and Chisholm (1977), Allen
(1978, 1979), Williams (1978), Allen and Williams (1978) and Ekes (1993) were not illustrated nor
adequately described and possibly correspond to additional examples of T. barretti. The burrow of
Edwards (1975) is insufficiently illustrated to determine confidently whether the specimen is walled
or unwalled.

Beaconites capronus (Howard and Frey, 1984) comb. nov.

1966 Chevron trail; Howard, p. 40, fig. 4.

1970 Chevron burrow; Frey and Howard, p. 183, fig. 8a [copy of Howard 1966, fig. 4],

1971 Chevron trail; Howard, p. 180, fig. 1 1 [jmrtim; copy of Howard 1966, fig. 4],

1972 Chevron trail; Howard, p. 217, fig. I [copy of Howard 1966, fig. 4],

1982 Chevron burrow; Frey and Howard, p. 3, fig. 2c [copy of Howard 1966, fig. 4],

*1984 Ancorichnus capronus Howard and Frey, p. 201, figs 2-3.

1984 Ancorichnus capronus Howard and Frey; Frey et al., p. 514.

1985 Ancorichnus capronus Howard and Frey; Frey and Howard, p. 373, figs 5.6, 5.8, 16.3d.

1985 Ancorichnus capronus Howard and Frey; Frey and Howard, p. 122, fig. 2 [copy of Howard and

Frey 1984, fig. 2],

1985 Ancorichnus capronus Howard and Frey; Frey and Pemberton, p. 90, fig. 26 [copy of Howard
and Frey 1984, fig. 2],

1989 Ancorichnus capronus Howard and Frey; Martino, p. 393, fig. 5.5.

1990 Ancorichnus capronus Howard and Frey; Frey, p. 204, fig. 7d.

1990 Ancorichnus capronus Howard and Frey; Frey and Howard, p. 808, fig. 12 [copy of Frey 1990,
fig. 7d], fig. 13. 1 .

Diagnosis. Thinly lined, smooth walled, predominantly horizontal cylindrical burrow with distinct
chevron-laminated fill (after Howard and Frey 1984).

Remarks. This ichnospecies is readily distinguishable by the chevron pattern of the meniscate fill.
This fill has frequently been described as comprising alternating coarse and fine chevron-shaped
packets of sediment (Howard and Frey 1984; Frey and Howard 1985u, 19856, 1990), although
alternating grain size is not necessarily ubiquitously present (e.g. Martino 1989). Howard (1966)
described a chevron trail preserved in convex epi relief that he interpreted as having been formed at
the sediment-water interface, probably by a gastropod. Subsequently the trace was considered to
be a burrowed structure (Frey and Howard 1970, 1982) and included within A. capronus by Howard
and Frey (1984). B. capronus has, to date, only been recognized in marine strata (Text-fig. 4).

320

PALAEONTOLOGY, VOLUME 37

Beaconites coronus (Frey et al ., 1984) comb. nov.

1968 Meniscus burrow; Daley, p. 124, fig. 12a \partim\.

1974 Horizontal burrows; Stanley and Fagerstrom, p. 70, figs 6, 7a-b.

1984 Muensteria Sternberg; Pemberton and Frey, p. 291, figs 4.6, 5b.

*1984 Ancorichnus coronus Frey et al ., p. 511, figs Id [copy of Stanley and Fagerstrom 1974, fig. 6], 1e,
3a-c.

Emended diagnosis. Predominantly horizontal, more rarely inclined to vertical, distinctly lined,
gently winding, small meniscate burrow. Relatively short (with respect to burrow width) meniscate
packets, or segments, of alternating sediment type. Menisci gently to moderately arcuate.

Remarks. This ichnospecies is distinguished from B. antarcticus by its more uniformly sized and thin
individual meniscate segments, and from B. capronus by the presence of gently to moderately
arcuate menisci. Typically, several segments occupy a length of the burrow equivalent to its overall
width. Only one of Daley’s (1968) meniscate trace fossils appears to retain evidence of a wall
structure. This structure, like the Taenidium of Toots (1967), may be due to preferential cementation
of the sediment adjacent to the trace-fossil boundary and not be a wall structure or part of the trace
fossil at all.

71833
71833
non 1833

1841

non 1851

1858

1858

non 1858
1861
non 1865

non 1869

1869

71877

71877

1877

mu

1877
71877
non 1880
non 1887

non 1887
1887

77 888
non 1888

Ichnogenus taenidium Fleer, 1887

Muensteria hoessii Sternberg, p. 33, pis 6.4, 7.3 [partim] [7= Taenidium isp., non Chondrites ].
Muensteria flagellar is Sternberg, p. 33, pi. 8.3 [indeterminate].

Muensteria geniculata Sternberg, p. 33, pi. 6.3 [= Hydrancylus isp. by Nathorst 1880;
7= Zoophycos isp.].

Muensteria schneideriana Goppert, p. 115, pi. 57.3 [7= Taenidium isp.].

Muensteria annulate Schafhautl, pp. 22, 140, pi. 8.9 [= T. fischeri according to Heer 1877;
7= Cladichnus isp.].

Muensteria ( Eumuensteria ) flagellar is Sternberg; Fischer-Ooster, p. 36.

Muensteria (Keckia) hoessii Sternberg; Fischer-Ooster, p. 38, pis 7.3, 16.4 [= Taenidium isp.],
16.5 [= Taenidium isp. partim ].

Muensteria (Keckia) dilatata Fischer-Ooster, p. 39, pi. 2 [7 = Zoophycos isp.].

Muensteria cretacea Ooster, p. 69, pi. 11.24 [reported in Heer 1877],

Muensteria annulate Schafhautl; Heer, p. 244, pi. 10.8 [= T. fischeri according to Heer 1877;
= Cladichnus isp.].

Muensteria Sternberg; Ooster, p. 29, pis 8. 2-8. 4 [8.2 = Spirophycus ; 8.3 7= Cladichnus fischeri'.
8.4 7 = Ophiomorpha ].

Muensteria Sternberg; Schimper, p. 194.

Muensteria ( Keckia ) antique Sternberg; Heer, p. 116, pi. 43.22.

Muensteria ( Eumuensteria ) flagellaris Sternberg; Heer, p. 116, pis 66.4-66.5.

Muensteria cretacea Ooster; Heer, p. 144, pi. 57.6 [= Taenidium isp.].

Muensteria ( Keckia ) nummulitica Heer, p. 163, pi. 69.4.

Muensteria ( Keckia ) hoessii Sternberg; Heer, p. 164, pis 66.6, 69.3 [= Taenidium isp.].
Muensteria (Keckia) dilatata Fischer-Ooster; Heer, p. 164.

Taenidium alysiodes Hosius and von der Marck, p. 131, pi. 24.5 [7 = alga],

Muensteria annulata Schafhautl; Squinabol, p. 554, pi. 17.3 [and synonymies therein;
= T. fischeri according to Squinabol 1891 ; 7 = Cladichnus fischeri],

Muensteria isseli Squinabol, p. 555, pi. 17.4-17.5 [= T. fischeri according to Squinabol 1891;
7 = Cladichnus fischeri],

Muensteria minima Squinabol, p. 555, pi. 16.5 [7 = Cladichnus isp.].

Taenidium carboniferum Sacco, p. 162, pi. 2.1.

Muensteria flagellaris Sternberg; Sacco, p. 169.

Muensteria annulata Schafhautl; Sacco, p. 170 [= T. fischeri according to Squinabol 1891;
= Cladichnus isp.].

KEIGHLEY AND PICKERILL BEACON ITES

321

1888
non 1888

?1890
non 1894
non 1941
1955
1958
1958
non 1962

1962

1964

71967

1971

71971

non 1971

1972

1974
non 1975

1975
1975
1975
1975

71976
1977
1977
1977
1977
1977
non 1977
1977
non 1977

71977
1977
non 1978

non 1978
1978
1978

1978
71978

1979

1979

1980
71982
71982

1983

1984
non 1984

1984
non 1984
71984

Muensteria minima Squinabol; Sacco, p. 170.

Muensteria isseli Squinabol; Sacco, p. 170 [= T. fischeri fide Squinabol 1891; = Cladichnus
fischeri ].

Taenidium helveticum Schimper and Schenk, p. 54, fig. 42.2.

Taenidium radiation Schroter, p. 80, figs 1-2 [= Cladichnus fischeri}.

Taenidium isseli Squinabol; Papp, p. 315, figs 1-2 [= Cladichnus fischeri],

Muensteria Sternberg; Seilacher, fig. 5.43.

Muensteria Sternberg; Seilacher, p. 1070, table 2.28.

Muensteria hoessii Sternberg; Seilacher, p. 1070. tabic 2.40 [= Taenidium isp.].

Taenidium Heer [parti in] ; Hantzschel, p. W218 [non figs 1 36.2a— 1 36.2b (copies of Papp, 1941 , fig.
1) = Cladichnus fischeri],

Muensteria hoessii Sternberg; Seilacher, p. 229, pi. 2.6 [= Taenidium isp.].

Muensteria Sternberg; Seilacher, p. 309, fig. 7.27.

Taenidium Heer; Toots, p. 93, fig. 1.

Taenidium Heer; Perkins and Stewart, p. 77, fig. 51d.

Taenidium Heer; Chamberlain, p. 42, fig. 6.32 [7 = Cladichnus isp., panim , 7 = Rhabdoglyphus
isp., partim ].

Taenidium annulata (Schafhautl); Chamberlain, p. 241, figs 8j-l, [7 = Cladichnus] pi. 32.12
[7 = Rhabdoglyphus ].

Taenidium Heer; Germs, p. 866, pis 2. 2-2. 3.

Muensteria Sternberg; Heinberg, p. 17, fig. I.

Munsteria [= lapsus calami ]; Chamberlain, p. 1076, fig. 2f'.

Taenidium Heer; Chamberlain, p. 1076, fig. 2g'.

Keckia Glocker; Hantzschel [partim], p. W75 [non fig. 47.2, copy of Glocker 1841. pi. 4],
Muensteria Sternberg; Hantzschel [partim], p. W84.

Taenidium Heer; Hantzschel, [partim], p. W112 [non fig. 70.1 = Cladichnus fischeri],
ITaenidium Heer; Hakes, p. 38, pi. I 1.6 [also tabulated in Hakes 1985]

Muensteria cf. M. hoessii Sternberg; Chamberlain, p. 14. figs 2l, 5e [7 = Taenidium barret ti],
Taenidium serpentium [lapsus calami]', Chamberlain, p. 18, fig. 3f [= Taenidium isp.].

Keckia annulata Glocker; Ksiazkiewicz, p. 63, pi. 3.14 [= Taenidium isp.].

Keckia hoessii (Sternberg); Ksiazkiewicz, p. 64, pis 3.15-3.16 [= Taenidium isps.].

Taenidium cumulation (Schafhautl); Ksiazkiewicz, p. 85, pi. 5.4 [= Taenidium isp.|

Taenidium isseli (Squinabol); Ksiazkiewicz. p. 85, pis 5. 1-5.2 [= Cladichnus isp |

Muensteria geniculata Sternberg; Ksigzkiewicz, p. 122, pi. 13.2 [7 = Taenidium barret ti],
Muensteria hamata Fischer-Ooster; Ksiazkiewicz, p. 122, pi. 13.3 [indeterminate branching
meniscate structure].

Muensteria planicostata Ksiazkiewicz, p. 122, pi. 13.1 [7 surface trail].

Taenidium Heer; Stanley et ah p. 267, fig. 18 [partim], fig. 19c.

Taenidium cumulation (Schafhautl); Alexandrescu and Brustur, p. 21, pi. 2.5 [= Cladichnus
fischeri],

Taenidium isp.; Alexandrescu and Brustur, p. 21, pi. 3.1 [= Cladichnus fischeri],

Muensteria Sternberg; Chamberlain, p. 144, fig. 3b.

Taenidium Heer; Chamberlain, p. 52, figs 4.23 [non 4.24 = Cladichnus], 10 [partim], I 1 [partim,
copy of Chamberlain 1971, fig. 6],

Taenidium Heer; Seilacher, p. 195, fig. 6.30.

Taenidium Heer; Carey, p. 438, fig. 6.

Muensteria Sternberg; Chamberlain p. 12, fig. 3 [copy of Chamberlain 1977, fig. 2],

Keckia isp. ; McCarthy, p. 363, figs 3a, 3c-e.

Taenidium Heer; Pickerill, p. 1270, fig. 4 f.

Taenidium Heer; MacDonald, p. 9, figs 6a-b.

Taenidium Heer; Tevesz and McCall, p. 270. fig. 5 [copy of Toots 1967, fig. 1].

Muensteria Sternberg; Wetzel, p. 290, fig. 2 [partim], fig. 6.4.

Muensteria Sternberg; Howard and Frey, p. 201, fig. 1.

Muensteria Sternberg; Pemberton and Frey, p. 291, figs 4.6, 5b [= Beaconites coronas].
Muensteria Sternberg; Pickerill et al„ p. 265, fig. 6a.

Muensteria Sternberg; Heinberg and Birkelund, p. 365, fig. 10b [= Ancorichnus cincorichnus}.
Backfilled burrows; Archer, p. 286, fig. 3c.

322

PALAEONTOLOGY, VOLUME 37

1985 Entradichnus meniscus Ekdale and Picard, p. 8, pi. 2a-b.

1985 Muensteria Sternberg; Eagar et al ., p. 140, pis Ib [partim\, 6c.

1985 Muensteria Sternberg; Frey and Pemberton, p. 76, fig. 2 [copy of Howard and Frey 1984, fig. 1],

1985 Muensteria Sternberg; Frey and Howard, p. 130, fig. 9 [copy of Howard and Frey 1984, fig. 1],

1985 Muensteria isp. ; Frey and Howard, p. 378, figs 10.12, 16. 3b, 19.6.

71986 Taenidium Heer; Miller, p. 343, fig. 3b.

1986 Muensteria isp.; Valenzuela et al ., p. 129.

1986 Muensteria Sternberg; Wheatcroft, p. 61, pi. 6.1c.

1986 Keckia isp.; Wheatcroft, p. 61, pi. 6.1b.

1986 Muensteria Sternberg; Brenchley et al., p. 246.

1987 Muensteria isp.; Narbonne et al., p. 1284.

non 1987 Muensteria isp.; Pickerill et a!., p. 83, fig. 4d [indeterminate].

1987 ‘ Muensteria ’ ; D’ Alessandro et al., p. 287, fig. 3.

1987 Muensteria isp.; Lockley et al., p. 259, figs 1, 2b.

71987 IMargariticlmus isp.; Lockley et al., p. 258, fig. 3.

1987 Backfilled burrow; Narbonne and Hofmann, p. 671, fig. lOe.

1988 Taenidium Heer; Bjerstedt, p. 55, fig. 5f.

1988 Muensteria isp.; Pickerill and Harland, p. 125, fig. 4 d.

1988 Muensteria geniculata Sternberg; McCann and Pickerill, p. 337, fig. 4.4 [= Taenidium isp.].
1988 Muensteria isp.; McCann and Pickerill, p. 337, fig. 4.5.
non 1988 Taenidium isseli Squinabol; McCann and Pickerill, p. 342, fig. 5.8 [7 = Cladichnus isp.].

1988 Muensteria Sternberg; Wiedman and Feldmann, p. 535, fig. 2.4.

1988 Beaconites antarcticus Vialov; Gordon, p. 144, figs 4a-b.

non 1989 Taenidium Heer; Powichrowski, p. 392, fig. 12 [= Cladichnus isp.].

1989 Muensteria isp.; Walter et al., p. 232, fig. 9a.

1990 Muensteria Sternberg; Dam and Andreasen. p. 215, fig. 11c.

1990 Muensteria cf. geniculata Sternberg; Mikulas, p. 314, fig. 1 a, pi. 2.4.

1991 Taenidium Heer; Burton and Link, p. 295, figs 7a, 7g.

1991 Taenidium Heer; Ekdale and Bromley, p. 232, figs 1, 3. 4a, 5, 7, 12.

71991 Taenidium Heer; Miller, p. 167, fig. 4h [copy of Miller 1986m, fig. 3b],

71991 Taenidium Heer; Miller, p. 76, fig. 4.

71991 Taenidium Heer; Romano, p. 197.

11991 Taenidium Heer; Scasso et al., p. 251.

1991 Muensteria Sternberg; Leszczyriski, p. 171, figs 4, 6 [parlim].

1992 Taenidium isp.; Crimes et al., p. 68, figs 4d, 5c.

71992 ITaenidium Heer; Mikulas, p. 225, pi. 2.3.

71993 Taenidium Heer; Smith et al., p. 590, fig. 13.

Emended diagnosis. Variably oriented, unwalled, straight, winding, curved, or sinuous, essentially
cylindrical, meniscate backfilled trace fossils. Secondary successive branching may occur, but true
branching is absent.

Type ichnospecies. T. serpentinum Heer, 1877.

Remarks. Wall linings are not present in Taenidium, distinguishing it as an unwalled ichnotaxon (see
previous discussion). The trace fossil may typically appear slightly annulate when thick meniscate
packets are present; otherwise, the boundary is usually slightly irregular. Fill is variable, of
homogeneous or heterogeneous, faecal and non-faecal content. Ichnospecies are defined by such
variations in the style of meniscate fill.

Only three valid ichnospecies of Taenidium were described by D'Alessandro and Bromley (1987),
T. serpentinum, T. satanassi and T. cameronensis. T. gillieroni Heer, 1877, and 7. convolutum Heer,
1877 were considered synonymous with the type. In this contribution we designate a fourth
ichnospecies, T. barretti.

D’Alessandro and Bromley (1987) also noted that M. clavata Sternberg, 1833, M. vermicu/aris
Sternberg, 1833 and M. lacunosa Sternberg, 1833 are not trace fossils. They also made T.fischeri the
type of a new ichnogenus, Cladichnus, included within which were specimens of T. lusitanicum Heer,

KEIGHLEY AND PICKERILL: BEACONITES

323

1881 by Heer (1881) and Wilkens (1947); M. (A.) annulata by Fiseher-Ooster (1858 partim)\
T. annulata by D’Alessandro et al. (1986) and ‘ Taenidium' by Kern ( 1978) and Pedersen and Surly k
(1983). Also included should be references of T. fischeri by Sacco (1888), Schimper and Schenk
(1890), Squinabol (1891), Schroter (1894) and Liburnau (1900). Together with M. bicornis Fleer,
1877, M. caprina Heer, 1877 and M. involutissima Sacco, 1888 that were placed within Spriophycus
Hantzschcl, 1962 by Hiintzschel (1962), M. clavata , M. vermicularis , M. lacunosa and T. fischeri are
not included in the above synonymy list.

A plethora of other potential synonyms also exist between the various ichnospecies of Taenidium
and Muensteria , and with other ichnogenera. Briefly, Fiseher-Ooster (1858) included Keckia
annulata Glocker, 1841, within one of his ‘subgenera' of Muensteria , namely M. ( Keckia ) annulata
Schafhautl, 1851. M. annulata and M. isseli Squinabol, 1887 were considered similar to T. fischeri,
by Heer (1877) and Squinabol (1891) respectively. Keckia and Saportia Squinabol, 1891, were
included within Wilcken’s (1947) discussion of Taenidium. Subsequently, D'Alessandro and
Bromley (1987) considered Keckia as a dubious ichnotaxon, and included Saportia within
Cladichnus. Most specimens of Keckia can therefore be accommodated within Cladichnus.

Another of Fischer-Ooster’s (1858) ‘subgenera’, containing mainly spreiten-like structures, was
Muensteria (Hydrancylus). The subgenus included M. (H.) geniculata Sternberg, 1833. Hydrancylus
was later (Nathorsl 1880) considered to be a distinct ichnogenus but can most probably be
considered a junior synonym of Zoophycos Massalongo, 1855. Ksi^zkiewicz’s (1977) interpretation
of Muensteria was generally analogous to Fischer-Ooster's ( 1858) Hydrancylus subgenus. However,
Ksiazkiewicz’s ( 1977) specimen of M. geniculata , as well as M. geniculata of McCann and Pickerill
( 1888) and Mikulas ( 1990), contained almost semi-circular menisci as opposed to a spreite, and must
be reassigned within Taenidium.

D’Alessandro and Bromley (1987) also considered M. cretacea Ooster, 1861, M. hoessii Heer,
1877, M. planicostata Ksiazkiewicz, 1977, T. carboniferum Sacco, 1888, and T. maeandriformis
M tiller, 1966 to be dubious ichnospecies. Of these, M. planicostata , in its original diagnosis, was
stated as hypichnial and described as ‘crescent grooves produced by contraction of (a gastropod)
foot, later cast in sand’ (Ksiazkiewicz 1977, p. 122). Being in all likelihood a surface trail, it should
not be included within Taenidium. T. meandriformis , like T. praecarbonicum Gtimbel, 1879 into
which the former was placed as a junior synonym by Pfeiffer (1966), is herein considered
synonymous with T. serpentinum. Sacco’s (1888) T. carboniferum is similar to the structures of
Lockley et al. (1987) that they erroneously termed IMargaritichnus Bandel, 1973; they should not
be included under Eione as suggested by Maples and Suttner (1990), Eione being invalid as an
ichnotaxon (Rindsberg 1990). In subdued convex epirelief or concave hyporelief, these trace fossils
resemble T. serpentinum , but although their structure is distinctly annulate, packets are occasionally
elliptical as opposed to bullet-shaped in form, and in longitudinal cross-section the upper parts of
the backfilling packets are imbricated. Imbrication and the occasional elliptical packeting (i.e. where
no menisci are really observed) may ultimately preclude inclusion of this form within Taenidium ,
and the ichnospecies is considered dubious. M. cretacea has characteristics similar to some
specimens of M. hoessii , and is similarly still considered a dubious ichnospecies. The original
M . hoessii and M.flagellaris Sternberg, 1833 were considered to resemble Chondrites by D'Alessandro
and Bromley (1987) because they were diagnosed as dichotomous. From Sternberg ( 1833, pis 6.4,
8.3), however, it is uncertain whether his specimens exhibited false or secondary successive
branching, and his other illustration of M. hoessii (pi. 7.3) is of an unbranched specimen. The
original diagnosis and illustrations of M. hoessii additionally mentioned the presence of ‘ lineis
transfer sis ’ and exhibit a distinct meniscate structure, unlike Chondrites.

M. hoessii is, however, a name that has been used for several different types of burrow and some
specimens may ultimately represent distinct ichnospecies (D’Alessandro and Bromley 1987).
Informally, it appears that the burrows M. cf. hoessii of Chamberlain (1977), (i)Taenidium of Carey
(1978) and Muensteria of Wiedman and Feldmann (1988) contain a highly distinct, heterogeneous
backfill in which regularly spaced segments form striking crescentic compartments between infill
that is indistinguishable from the enclosing strata. K. annulata and K. hoessii of Ksiazkiewicz (1977,

324

PALAEONTOLOGY, VOLUME 37

pis 3.14, 3.16) and Keckia isp. of McCarthy (1979, fig. 3c) are similar to B. capronus apart from
lacking a wall. Longitudinally thin, equally spaced, deeply concave, densely stacked but very
distinct heterogeneous meniscate segments would seem to distinguish M. ( K .) Iioessii of Fischer-
Ooster (1858, pi. 16.4), M. cretacea of Fleer (1877), M. Iioessii ( partim ) of Fleer (1877), and
M. Iioessii of Seilacher (1958, 1962). We therefore concur with D’Alessandro and Bromley (1987)
that additional ichnospecies of Taenidium can probably be distinguished. Apart from T. barretti
introduced herein, however, we have not attempted to investigate thoroughly other potential
ichnospecies of Tctenidium , as this is beyond the scope of this contribution. Such material has been
retained in the ichnogeneric synonymy only.

Ekdale and Bromley (1991) also illustrated both the densely meniscate and strikingly segmented
forms of Taenidium occurring in association with Zoopliycos. Their photographs additionally
illustrate the problems that can arise between these two ichnotaxa when seen in section, because
both may appear as unwalled, meniscate (or spreiten) structures. Slabbing of samples may be
necessary to distinguish the two (Chamberlain 1978<7). The ichnogenus Compaginatichnus Pickerill,
1989, was erected for burrows that contain similar arcuate meniscate infill to Taenidium. However,
at the base of this burrow, underlying the meniscate infill, is a distinct coprolitic, pelletal layer. The
burrow therefore contains a compound fill of upper meniscate segments and lower pelletal layer.
Recognition of both types of fill is necessary for the correct identification of this ichnotaxon;
otherwise, if the burrow was entirely meniscate, it would be included within Taenidium (Pickerill
1989). Full relief views Arthrophycus Hall, 1852, and Planolites annularius Walcott, 1980, may also
give the impression of being meniscate and could be confused with Taenidium , as could Nereites
MacLeay, 1839, where the disturbance zone in the latter is not preserved (D’Alessandro and
Bromley 1987). Again, slabbing of samples is essential to confirm that the trace fossil being
identified is a burrow and not a trail, and that the infill is entirely meniscate, and not partly pelletal,
internally structureless or imbricate, nor externally disturbing the surrounding sediment.
Imponoglyp/ius Vialov, 1971, contains truncated cones invaginated into one another. Poorly
preserved or weathered material may potentially appear as simple, articulated meniscate segments
assignable to Taenidium. Another invaginate trace fossil, Rhabdoglyphus Vassoievich, 1951.
resembles material figured by Chamberlain (1971c/, pi. 32. 12) as unbranched T. anmdata (see Stanley
and Pickerill 1993 for discussion).

Taenidium serpentinum Heer, 1877

1858 Muensteria ( Keckia ) schneideriana Goppert; Fischer-Ooster, p. 39, pi. 15.3.

*1877 Taenidium serpentinum Heer, p. 116, pis 46.3-46.4.

1877 Taenidium gillieroni Heer, p. 117, pi. 50.1 [partim],

1877 Taenidium convolution Heer, p. 117, pi. 50.2 [partim],

1887 Muensteria serpentina [= lapsus calami]: Maillard, p. 37, pi. 1.4.

1890 Taenidium serpentinum Heer; Schimper and Schenk, p. 54, fig. 42.1.

71966 Taenidium maeandriformis Muller, p. 712, figs 1 -2, pi. 1.

71966 Taenidium praecarbonicum Giimbel; Pfeiffer, p. 688, fig. 3.21.

71971 Taenidium serpentium [= lapsus calami ]; Chamberlain, p. 42, fig. 6.32 [partim],

1971 Taenidium serpentium [= lapsus calami ]; Chamberlain, p. 241, pi. 32.10.

71972 Taenidium carbonicum [= lapsus calami ]; Hantzschel, p. 115, fig. 3.

1974 Muensteria Sternberg; Fiirsich. p. 34. figs 28, 29 a.

71977 Taenidium serpentium [= lapsus calami ]; Chamberlain, p. 18, figs 2a, 2 /;, 3f 7a [non 3i =
Taenidium isp.].

71982 Taenidium praecarbonicum Giimbel; Benton, p. 122. fig. 6r.

1987 Taenidium serpentinum Heer; D’Alessandro and Bromley, p. 743, figs 5-7.

71989 Taenidium serpentinum Heer; Miller, p. 48, fig. 2b.

1990 Taenidium serpentinum Heer; Dam, p. 142, fig. 11 a [partim],

71990 Taenidium serpentinum Heer; Maples and Suttner, p. 874, fig. 14.1.

1990 Taenidium serpentinum Heer; Dam, p. 226, fig. 6 [partim copy of Dam, 1990, fig. I In], figs 15,
19, 23.

KEIGHLEY AND P1CKERI LL: BEACON ITES

325

Diagnosis. Serpentiform Taenidium having well-spaced, arcuate menisci; distance between menisci
about equal to or a little less than burrow width. External moulds may show slight annulation
corresponding to menisci, or fine transverse wrinkling. Secondary subsequent branching and
intersections occur. Boundary sharp and lacks lining (after D' Alessandro and Bromley 1987).

Remarks. As with T. satanassi, this ichnospecies has meniscate packets typically of slightly less
length than width. Unlike T. satanassi , however, the fill is homogeneous and of the same
composition as the enclosing strata. The possibility of a thin, incomplete, and generally
‘insignificant' lining occurring locally was indicated by D’Alessandro and Bromley (1987). This
should not affect the naming of the structure, since trace fossils are named after their predominant
features (Pickerill 1994), in this case the lack of a wall structure. T. praecarbonicum is provisionally
included within T. serpentinum (along with specimens of T. gilleroni , and T. convolution ), although
sketches of the former (Heer 1877; Pfeiffer 1966; Benton 1982) would indicate that assignment is
dubious, since the individual sediment packets are ellipsoidal in shape and nowhere have concave
meniscate interfaces. To date, the ichnospecies has only been confirmed in marine sediments (Text-
fig. 5).

Taenidium satanassi D’Alessandro and Bromley, 1987

Muensteria isp. Sternberg; D’Alessandro et a/., p. 299, fig. 5b.

Taenidium satanassi D’Alessandro and Bromley, p. 743, figs 6, 8-9.

Taenidium satanassi D’Alessandro and Bromley; Bromley, p. 178, fig. 10.9 [copy of
D’Alessandro and Bromley 1987, fig. 8].

Taenidium satanassi D’Alessandro and Bromley; Frey and Howard, p. 15, fig. 16.13, fig. 25
[copy of Howard and Frey 1984, fig. 1],

Diagnosis. Sinuous to nearly straight backfilled burrows, the fill consisting of meniscate packets,
each packet containing two types of sediment of more or less equal thickness; sediment packets
considerably shorter than wide. Menisci weakly arcuate (after D’Alessandro and Bromley 1987).

Remarks. Typically a slight constricting annulation in the outer boundary of the burrow
corresponds to the contact between the packets. T. barretti , although similarly containing a
heterogeneous fill, does not have distinct packeting but only short, more arcuate crescents or non-
compartmentalized meniscate fill. Frey and Howard’s (1990) specimens do not incorporate
alternating meniscate fill, although individual packets are considerably shorter than wide. The few
specimens so far encountered of this ichnospecies are from a marine setting (Text-fig. 5).

1986

*1987

1990

1990

Taenidium cameronensis (Brady, 1947)

*1947 Scolecocopnts cameronensis Brady, p. 471, pi. 69, fig. 1.

1955 Scolecocopnts cameronensis Brady; Lessertisseur, p. 58, fig. 33c [copy of Brady, pi. 69].

1978 Scolecocoprus cameronensis Brady; Decourten, p. 491, fig. Ia-c.

1987 Taenidium cameronensis (Brady); D’Alessandro and Bromley, p. 743, fig. 6.

1993 Taenidium cameronensis (Brady); Pickerill et al ., p. 63, fig. 2d.

Diagnosis. Unwalled meniscate burrows, secondary successive branching and intersection may be
present. Meniscate packets usually longer than wide, with the deeply concave meniscate interfaces
resulting in a nested appearance. (After D’Alessandro and Bromley 1987).

Remarks. This ichnotaxon was not clearly redefined by D’Alessandro and Bromley (1987).
Although they retained the overall meaning of Scolecocoprus Brady, 1947, the accompanying
illustration of the ichnotaxon by D’Alessandro and Bromley (1987, fig. 6) cannot be distinguished
from the original illustration of Scolecocoprus arizonensis Brady, 1947. S. arizonensis has a different

326

PALAEONTOLOGY, VOLUME 37

LOCATIONS *aulhors

AGE

ENVIRONMENT

-5u-e

s »

3 .8
> >

5 $ && w, Id

®1 £2
1 i S 5
I <5 li le

I

o

Is

IS

S Italy

2

Namibia

S California, USA

Utah, USA4
5.6

S Italy

Montana, USA
8

Jamaica
Utah, USA

to

Central Portugal
S England

12

Switzerland

13.14

Jameson Land, Greenland
15,16

S India

17,18

NW Argentina
SW England

20.21

Arizona, USA

22-24

Oklahoma, USA
25,26

Ireland

27

N England

2829

Nova Scotia, Canada

29,30

New Brunswick, Canada
31,32

Central Scotland
32

S Wales

33-36

Antarctica

37.38

New York State, USA
39

S Norway

40

New York State, USA

Pleistocene

Pleistocene

Miocene

Eocene

Eocene

Palaeocene

Palaeogene

Cretaceous

Jurassic

Jurassic

Jurassic

Jurassic

Triassic

Permian

Permian

Permian

Carboniferous

Carboniferous

Carboniferous

Carboniferous

Carb+ Devonian

Carb/Devonian

Devonian

Devonian

Devonian

Silurian

Ordovician

KEY: ♦ » T. serpentinum m—mm T.satanassi ^ T. cameronensis ■■■■ T. barretti

text-fig. 5. Temporal and environmental distribution of Taenidium ichnospecies. Only references to
confidently assigned ichnospecies and of known environment are included. Authors: 1, D’ Alessandro et al.
1993; 2, Smith et al. 1993; 3, Squires and Advocate 1984; 4, D'Alessandro et al. 1987; 5, D'Alessandro et al.
1986; 6, D’Alessandro and Bromley 1987; 7, Diemer and Belt 1991 ; 8, Pickerill et al. 1993; 9, Bracken and
Picard 1984; 10, Fursich 1981 ; 1 1, Fursich 1974; 12, Heer 1877; 13, Dam 1990r/; 14, Dam 19906; 15, Maulik
and Chaudhuri 1983; 16, Sarkar and Chaudhuri 1992; 17, Acenolaza and Buatois 1991; 18, Acenolaza and
Buatois 1993; 19, Ridgway 1974; 20, Brady 1947; 21, Decourten 1978; 22, Chamberlain 1971a; 23,
Chamberlain 19716; 24, Chamberlain 19786; 25, Graham and Pollard 1982; 26, Briick 1987; 27, Eagar et al.
1985; 28, Keighley and Pickerill 1993; 29, this paper; 30, Nilsen 1982; 31, Allen and Williams 19816; 32,
Pearson 1992; 33, Gevers et al. 1971 ; 34. Bradshaw 1981 ; 35, Bradshaw and Webers 1988; 36, Woolfe 1990;
37, Thoms and Berg 1985; 38, Bridge et al. 1986; 39, Dam and Andreasen 1990; 40, Tegan and Curran 1992.

KEIGHLEY AND PICKERILL: BEACONITES

327

text-fig. 6. Taenidium barretti (Bradshaw). Division of Natural Sciences, New Brunswick Museum, Saint John
(NBMG); Grand Etang, Cape Breton Island, Nova Scotia, Canada; Pomquet Formation, middle
Carboniferous; burrows from fluvio-lacustrine (shoreline) deposits, a i, ii, NBMG 9074; stereopair of full relief
structures; thin-sectioning of the structures confirms that they are unwalled, the knobbly outline being the
result of irregular meniscate backpacking of sediment containing angular mudstone fragments; x0-7.
B, NBMG 9217; top surface view of horizontally orientated (and inclined) specimens that have more uniform,
silt-grade backfill, resulting in very indistinct menisci being preserved; x 0-25. c, NBMG 9216; vertical section
through slab, showing irregular meniscate structure that, in this case, indicates downward movement of

producer; xl.

type of meniscate fill, approaching a chevron-shape, and is not synonymous with Taenidium
because this ichnospecies contains a deeply grooved ornamentation at the base of the burrow.
T. cameronensis , however, remains distinguishable from the similarly homogeneously backfilled
T. serpentinum primarily by having packets longer than the burrow width.

Taenidium barretti (Bradshaw, 1981)

Plate 1 ; Text-figure 6

1968 ‘ Scolicia ' de Quatrafages; Webby, p. 1003, fig. 8.

328

PALAEONTOLOGY, VOLUME 37

1971 Beaconites antarcticus Vialov; Gevers et al., p. 81, figs 1-2, 4.

1974 Problematica [= cf. Beaconites antarcticus Vialov; Pollard, 1976]; Ridgway, p. 511, fig. l,pl. 17.

1975 Beaconites Vialov; Hantzschel, p. W45, fig. 28.1 [copy of Gevers et al., 1971, fig. 2],

1981 Beaconites antarcticus Vialov; Allen and Williams, p. 255, figs 6-11.

*1981 Beaconites barretti Bradshaw, p. 630, figs 17-18.

1981 Scoyenia isp.; Fursich, p. 160, pi. 5 \partim\.

1982 Beaconites antarcticus Vialov; Graham and Pollard, p. 259, figs 3a, 4a-c, 5a-b.

1982 Backfilled burrow; Nilsen, p. 79, fig. 46a.

71982 Meniscate burrow; Bown, p. 282, fig. 12b.

71982 Meniscate burrow; Bown and Kraus, p. 118, figs 7e-f, 8a-b.

1983 Horizontal feeding burrow; Maulik and Chaudhuri, p. 23, fig. 3.

1984 Muensteria isp.; Braken and Picard, p. 482, fig. 9.

1984 7 Muensteria isp.; Squires and Advocate, p. 594, figs 2a-f.

1985 cf. Beaconites antarcticus Vialov; Eagar et al., p. 134, pi. 14a.

1985 Bivalve trace fossils; Thoms and Berg, p. 13, pi. lc-E,

1986 Vertical burrows; Bridge et al., p. 65, pi. 1b.

1987 Beaconites barret ti Bradshaw; p. 259, figs 4-5.

1987 cf. Ancorichnus coronas Frey et al. ', D’ Alessandro et al., p. 285, fig. 2.

1988 Beaconites barretti Bradshaw; Bradshaw and Webers, p. 787.

1990 Beaconites isp; Dam and Andreasen, p. 208, fig. 11b.

1990 Beaconites barretti Bradshaw; Woolfe, p. 302, fig. 3.

1991 Ancorichnus coronus Frey et al.; Acenolaza and Buatois, p. 96, pi. 2.1.

1991 Meniscate burrow; Diemer and Belt, p. 97, fig. 12 [non Rhizocorallium, p. 96].

1992 7 Beaconites Vialov; Pearson, p. 129, fig 3d.

1992 Beaconites barretti Bradshaw; Tegan and Curran, p. A154.

1992 Taenidium Heer; Sarkar and Chaudhuri, p. 1 1, figs 4-5.

1993 Taenidium isp.; D'Alessandro et al., p. 497, figs 3, 4b, 6b, 9, 10b [partim] 2, 6a, 10a, 12b.

1993 Taenidium Heer; Smith et al., p. 590, fig. 14.

1993 Ancorichnus coronus Frey et a/. ; Acenolaza and Buatois, p. 188, fig. 4d [copy of Acenolaza and
Buatois 1991, pi. 2.1].

1993 Taenidium isp.; Keighley and Pickerill. p. 83.

1993 Taenidium barretti (Bradshaw); Keighley and Pickerill, p. 83.

Emended diagnosis. Straight to variably meandering, unbranched, unwalled, meniscate backfilled
burrow. Menisci are commonly hemispherical or deeply arcuate, tightly packed or stacked, forming
non-compartmentalized backfill or thin meniscate segments.

Remarks. As previously discussed, menisci may merge laterally at the burrow boundary and in some
preservational variants form a pseudo-wall or -lining. This is more likely where distinct segmentation
of the backfill has been achieved. The boundary may be irregular to crenate, with individual
meniscate segments slightly offset to one another (Graham and Pollard 1982). In full relief the
burrow boundary may appear knobbly and similar to Ophiomorpha irregulaire Frey et al., 1978
(Text-fig 6a). The distinctiveness of menisci in the backfill is variable (Plate 1; Text-fig. 6b-c).
Individual backfilled compartments may be so short in longitudinal section (thin segments) with
respect to overall width, that distinct segmentation of the backfill is not achieved and the fill
becomes irregular (non-compartmentalized). In addition, homogeneity in particle size or clast
composition may result in the menisci being poorly defined when produced, and result in them being
uniformly cemented and weathered. Many specimens from the Upper Palaeozoic are of giant size
(up to 450 mm wide - Pearson 1992), although recorded widths (but not necessarily diameters -
Graham and Pollard 1982) may be as small as 5 mm (Plate 1 ; Text-fig. 3). Almost all recordings are
from non-marine environments (Text-fig. 5).

T. barretti is typically undulating and subparallel to stratification, although vertical sections, of
similar diameter as the (sub-) horizontal burrow, may be dominant (e.g. Allen and Williams 1981a,
Graham and Pollard 1982, Bracken and Picard 1984). Although previously interpreted as ‘escape’
or ‘equilibrium’ structures, some exclusively vertical ‘burrows’ are also included in the synonymy.

KEIGHLEY AND PICKERILL: BEACONITES

329

Simple downward displacement of primary sedimentary laminae is not an exclusive and
distinguishing feature that can separate equilibrium structures from ichnospecies of repichnial
Taenidium , that comprise arcuate non-faecal backfill (see Sarkar and Chaudhuri 1992, fig. 5).

CONCLUSIONS

There is still a tendency for ichnologists to be influenced in naming a particular trace fossil
depending on the interpreted depositional environment in which it is encountered. The classic
example of this is the Cruziana-Rusophycus Hall, 1852 versus Isopodichnus Bornemann, 1899
debate, whereby several authors still persist in using the latter if a bilobate trace is encountered in
non-marine rocks (see Bromley 1990). Similarly, the presence of large meniscate burrows in non-
marine to marginal marine deposits appears to have automatically resulted in immediate
comparison to Beaconites and, more recently, small lined backfilled structures to Ancorichnus
without careful consideration of the significant criteria necessary for their nomenclature.
Ancorichnus is unwalled but possesses a two zoned fill : an outer mantle with an internal, transverse
to diagonal fabric, and an inner meniscate backfill. Simple meniscate structures with a distinct but
unornamented and unstructured wall are considered to belong to Beaconites , whose ichnospecies
are differentiated based on variations within the meniscate infill. Taenidium is a simple, unwalled,
meniscate, backfilled structure, ichnospecies again being differentiated on the basis of variation
within the meniscate infill (Text-fig. 2). We re-emphasize that depositional environment is not a
valid diagnostic criterion, and workers should name their trace fossils solely on the basis of
morphology. Without this directive, the development of workable ichnofacies/ichnocoenoses
cannot be continued, since workers will be following the circular argument whereby a trace fossil
is being utilized as an aid in the interpretation of a specific palaeoenvironment, but that
palaeoenvironment is being inferred in the first place to name the trace fossil.

Acknowledgements. We thank R. MacNaughton and G. Narbonne for valuable discussion, C. Stanley for
considered reviews of several versions of this manuscript, and R. Bromley and C. Cleal for suggesting
important modifications, C. Colborne, A. Gomez, R. McCulloch, A. Murphy, D. Pirie and D. Tabor for
technical assistance, and J. Kraus for providing translations of German manuscripts. The work was funded by
an NSERC Canada grant to R.K.P, and constitutes part of D.G. K.’s Ph D thesis.

REFERENCES

acenolaza, f. G. and buatois, l. a. 1991. Trazas fosiles del Paleozoico superior continental Argentino.
Ameghiniana , 28, 89-108.

- 1993. Nonmarine perigondwanic trace fossils from the late Paleozoic of Argentina. Ichnos , 2.
183-203.

alexandrescu, G. and brustur, t. 1978. Asupra unor urme de activiate organica (trace fossils) din fli§ul
Carpafilor orientali (Partea I). Dari de Seamd ale Sedintelor. Institut de Geologie si Geofizica, 65, 17-30.
allen, j. r. l. 1961 . Sandstone-plugged pipes in the Lower Old Red Sandstone of Shropshire, England. Journal
of Sedimentary Petrology, 31, 325-335.

1963. Depositional features of Dittonian rocks: Pembrokeshire, compared with the Welsh Borderland.
Geological Magazine, 100, 385-400.

1978. Locality B15 - Gelliswick Bay, Dyfed. 87-88. In friend, p. f. and williams, b. p. j. (eds). A field
guide to selected outcrop areas of the Devonian of Scotland , the Welsh Borderland, and South Wales.
Palaeontological Association, London, 106 pp.

— 1979. Old Red Sandstone facies in external basins, with particular reference to southern Britain. 65-80.
In house, m. r., scrutton, c. T. and bassett, m. G. (eds). The Devonian System. Special Papers in
Palaeontology, 23, 1-353.

— and williams, b. p. j. 1978. The sequence of the earlier Lower Old Red Sandstone (Siluro-Devonian),
north of Milford Haven, southwest Dyfed (Wales). Geological Journal, 13, 113-136.

1981a. Beaconites antarcticus : a giant channel-dwelling trace fossil from the Lower Old Red
Sandstone of South Wales and the Welsh Borders. Geological Journal, 16, 255-269.

330

PALAEONTOLOGY, VOLUME 37

allen, J. r. l. and williams, b. p. j. 1981/). Sedimentology and stratigraphy of the Townsend Tuff Bed (Lower
Old Red Sandstone) in South Wales and the Welsh Borders. Journal of the Geological Society , London , 138,
15-29.

archer, a. w. 1984. Preservational control of trace-fossil assemblages: Middle Mississippian carbonates of
south-central Indiana. Journal of Paleontology, 58, 285—297.

BAMFORD, M. L., BRUCK, P. M., COOPER, M. A., FORBES, W. H. and MACCARTHY, I. A. J. 1986. Beaconites - type
burrows from the Old Red Sandstone of Hook Head, Co. Wexford, Ireland. Proceedings of the Geologists'
Association , 97, 59—7 1

bandel, K. 1973. A new name for the ichnogenus Cylindrichnus Bandel, 1967. Journal of Paleontology, 47, 1002.
benton, M. J. 1982. Dictyodora and associated trace fossils from the Palaeozoic of Thuringia. Lethaia , 15,
115-132.

berg, t. m. 1977. Bivalve burrow structures in the Bcllvale sandstone. New Jersey and New York. Bulletin of
the New Jersey Academy of Science, 22, 1-5.

bjerstedt, t. w. 1988. Multivariate analyses of trace fossil distribution from an Early Mississippian oxygen-
deficient basin. Central Appalachians. Palaios, 3, 53-68.
bornemann, j. g. 1899. Uber den Buntsandstein in Deutschland seie Bedeutung fur die Trias. Beitrdge zur
Geologie und Paldontologie, 1, 1-61.

bown, t. m. 1982. Ichnofossils and rhizoliths of the nearshore fluvial Jebel Qatrani Formation (Oligocene),
Fayum Province, Egypt. Palaeogeography, Palaeoclimatology, Palaeoecology, 40, 255-309.

— and kraus, m. j. 1983. Ichnofossils of the alluvial Willwood Formation (Lower Eocene), Bighorn Basin,
Northwest Wyoming, U.S.A. Palaeogeography, Palaeoclimatology, Palaeoecology, 43, 95-128.

bracken, b. and picard, m. d. 1984. Trace fossils from Cretaceous/Tertiary North Horn Formation in central
Utah. Journal of Paleontology, 58, 477-487.

bradshaw, m. a. 1981 . Palaeoenvironmental interpretations and systematics of Devonian trace fossils from the
Taylor Group (Lower Beacon Supergroup), Antarctica. New Zealand Journal of Geology and Geophysics, 24,
615-652.

— and webers, G. f. 1988. The Devonian rocks of Antarctica. 783-795. In mcmillan, n. j., embry, a. f. and
glass, d. j. (eds). Devonian of the World. 1 Proceedings of the Second International Symposium on the
Devonian System, Calgary. Canadian Society of Petroleum Geologists, 795 pp.

brady, l. f. 1947. Invertebrate tracks from the Coconino Sandstone of northern Arizona. Journal of
Paleontology, 21, 466-472.

brenchley, p. j., romano, m. and Gutierrez- marco, j. c. 1986. Proximal and distal hummocky cross-stratified
facies on a wide Ordovician shelf in Iberia. 241-255. In knight, r. j. and mcClean, j. r. (eds). Shelf sands
and sandstones. Memoirs of the Canadian Society of Petroleum Geologists, 11, 1-347.
bridge, j. s. and droser, m. L. 1985. Unusual marginal-marine lithofacies from the Upper Devonian Catskill
clastic Wedge. 143-161. In woodrow, o. l. and sevon, w. d. (eds). The Catskill Delta, Geological Society
of America, Special Paper, 201, 1-246.

— and Gordon, e. a. 1985. Quantitative interpretation of ancient river systems in the Oneonta Formation,
Catskill Magnafacies. 168-181. In woodrow, o. l. and sevon, w. d. (eds). The Catskill Delta. Geological
Society of America, Special Paper, 201, 1-246.

— and titus, R. c. 1986. Non-marine bivalves and associated burrows in the Catskill Magnafacies, New
York State. Palaeogeography, Palaeoclimatology, Palaeoecology, 55, 65-77.

— and nickelsen, b. h. 1985. Reanalysis of the Twilight Park Conglomerate, Upper Devonian Catskill
Magnafacies, New York State. Northeastern Geology, 1, 181-191.

— van veen, p. m. and matten, l. c. 1980. Aspects of the sedimentology, palynology, and palaeobotany of
the Upper Devonian of Southern Kerry Head, Co. Kerry, Ireland. Geological Journal, 15, 143-170.

bromley, R. G. 1990. Trace fossils: biology and taphonomy. Special Topics in Palaeontology, 3, 1-280.
bruck, p. m. 1987. A note on the trace fossil Beaconites barretti in the Old Red Sandstone of Co. Dublin,
Ireland. Proceedings of the Geologists' Association, 98, 259-263.

— forbes, w. h., nance, D. and pickerill, R. K. 1985. Beaconites antarcticus in the (?middle) Late Devonian
McAras Brook Formation, Cape George, Nova Scotia. Maritime Sediments and Atlantic Geology, 21, 87-96.

burton, b. r. and link, p. k. 1991. Ichnology of fine-grained mixed carbonate-siliclastic turbidites. Wood
River Formation, Pennsylvanian-Permian, south-central Idaho. Palaios, 6. 291-301.
carey, J. 1978. Sedimentary environments and trace fossils of the Permian Snapper Point Formation, southern
Sydney basin. Journal of the Geological Society of Australia , 25, 433-458.
chamberlain, c. k. 1971 u. Morphology and ethology of trace fossils from the Ouachita Mountains, southeast
Oklahoma. Journal of Paleontology, 45, 212-246.

KEIGHLEY AND PICKERILL: BEACONITES

331

— 1971 b. Bathymetry and paleoecology of Ouachita Geosyncline of southeastern Oklahoma as determined
from trace fossils. Bulletin of the American Association of Petroleum Geologists , 55, 34—50.

— 1975. Trace fossils in DSDP cores of the Pacific. Journal of Paleontology, 49. 1074-1096.

— 1977. Ordovician and Devonian trace fossils from Nevada. Bulletin of the Nevada Bureau of Mines and
Geology , 90, 1-24.

1978a. Recognition of trace fossils in cores. 119-166. In basan, p. b. (ed . ). Trace fossil concepts. Society
of Economic Paleontologists and Mineralogists , Short Course , 5, 1 181.

— 1978fi. A guidebook to the trace fossils and paleoecology of the Ouachita geosyncline. Society of Economic
Paleontologists and Mineralogists, 68 pp.

— 1979. Trace-fossil biofacies in the Lower and Middle Paleozoic of central Nevada. Bringham Young
University , Geology Studies , 26, 9-19.

clifton, h. e. and Thompson, j. k. 1978. Macaronichnus segregatis: a feeding structure of shallow marine
polychaetes. Journal of Sedimentary Petrology , 48, 1293-1302.
crimes, t. p., garcia hidalgo, j. F. and poire, d. g. 1992. Trace fossils from Arenig flysch sediments of Eire
and their bearing on the early colonisation of the deep seas. Ichnos , 2, 61-77.
d’alessandro, a. and bromley, r. g. 1987. Meniscate trace fossils and the Muensteria-Taenidium problem.
Palaeontology , 30, 743-763.

— ekdale, a. a. and sonnino, m. 1986. Sedimentologic significance of turbidite ichnofacies in the Saraceno
Formation (Eocene), southern Italy. Journal of Sedimentary Petrology, 56, 294—306.

— and picard, M. d. 1987. Trace fossils in fluvial deposits of the Duchesne River Formation (Eocene),
Uinta Basin, Utah. Palaeogeography, Palaeoclimatology, Palaeoecology, 61, 285-301.

— loicano, f. and bromley, r. g. 1993. Marine and nonmarine trace fossils and plant roots in a regressional
setting (Pleistocene, Italy). Rivista Italiana di Paleontologia e Stratigrafia, 98, 495-522.

daley, b. 1968. Sedimentary structures from a non-marine horizon in the Bembridge Marls (Oligocene) of the
Isle of Wight, Hampshire, England. Journal of Sedimentary Petrology , 38, I 14—127.
dam, G. 1990m Taxonomy of trace fossils from the shallow marine Lower Jurassic Neill Klinter Formation,
East Greenland. Bulletin of the Geological Society of Denmark , 38, 1 19-144.

— 1990fi. Palaeoenvironmental significance of trace fossils from the shallow marine Lower Jurassic Neill
Klinter Formation, East Greenland. Palaeogeography , Palaeoclimatology , Palaeoecology , 79, 221-248.

and andreasen, f. 1990. High-energy ephemeral stream deltas; an example from the Upper Silurian

Holmestrand Formation of the Oslo Region, Norway. Sedimentary Geology, 66, 197-225.
decourten, f. l. 1978. Scolecocoprus cameronensis Brady (1947) from the Kaibab Limestone of the Northern
Arizona: a reinterpretation. Journal of Paleontology, 52, 491-493.
diemer, j. a. and belt, e. s. 1991. Sedimentology and paleohydraulics of the meandering river systems of the
Fort Union Formation, southeastern Montana. Sedimentary Geology, 75, 85-108.
dineley, d. l. 1966. The Dartmouth Beds of Bigbury Bay, South Devon. Journal of the Geological Society ,
London, 122. 187-217.

dixon, E. e. l. 1921. Geology of the South Wales coalfield. Part XIII. The country around Pembroke and
Tenby. Memoirs of the Geological Survey of Great Britain, 1-220.
eagar, r. m. c., baines, J. G., collinson, j. d., hardy, p. G., okolo, s. a. and pollard, j. e 1985. Trace fossil
assemblages and their occurrence in Silesian (mid-Carboniferous) deltaic sediments of the central Pennine
Basin, England, 99-149. In curran, h. a. (ed.). Biogenic structures: their use in interpreting depositional
environments. Special Publications, Society of Economic Paleontologists and Mineralogists, 35, 1-347.
edwards, p. 1975. Invertebrate burrows in an Oligocene fresh-water limestone. Contributions to Geology, 14,
7-8.

ekdale, a. a. and picard, m. d. 1985. Trace fossils in a Jurassic eolianite, Entrada Sandstone, Utah, U.S.A.
3-12. In curran, h. a. (ed.). Biogenic structures: their use in interpreting depositional environments. Special
Publications Society of Economic Paleontologists and Mineralogists, 35, 1-347.

— and bromley, r. g. 1991. Analysis of composite ichnofabncs: an example in uppermost Cretaceous chalk
of Denmark. Palaois, 6, 232-249.

ekes, c. 1993. Bedload-transported pedogenic mud aggregates in the Lower Old Red Sandstone in southwest
Wales. Journal of the Geological Society, London, 150, 469-471.
fischer-ooster, c. 1858. Die fossilen Fucoiden der Schweizer-Alpen, nebst Erorterungen iiber deren geologisches
Alter. Huber, Bern, 72 pp.

FORSYTH, i. h. and CHISHOLM, j. I. 1977. The geology of East Fife. Memoirs of the Geological Survey of Great
Britain, 1-284.

332

PALAEONTOLOGY, VOLUME 37

frey, r. n. 1990. Trace fossils and hummocky cross-stratification. Upper Cretaceous of Utah. Palaois , 5,
203-218.

and Howard, j. d. 1970. Comparison of Upper Cretaceous ichnofaunas from siliceous sandstones and
chalk. Western Interior Region, U.S.A. 141-166. In crimes, t. p. and harper, j. c. (eds). Trace fossils.
Geological Journal Special Issue , 3, 1-547.

1982. Trace fossils from the Upper Cretaceous of the Western Interior. Mountain Geologist , 19. 1-10.
1 985<3. Trace fossils from the Panther member. Star Point Formation (Upper Cretaceous), Coal
Creek Canyon, Utah. Journal of Paleontology , 59, 370-404.

— 1985fi. Upper Cretaceous trace fossils. Book Clifts of Utah: a field guide. 115-152. In kamola,
d. l., pfaff, b. j., newman, s. l., frey, r. n. and Howard, j. d. (eds). Depositional facies of the Castlegate
and Blackhawk Formations, Book Cliffs, eastern Utah. Society of Economic Paleontologists and
Mineralogists midyear meeting , Golden , Colorado , Field Trip 70, 152 pp.

1990. Trace fossils and depositional sequences in a clastic shelf setting. Upper Cretaceous of Utah.
Journal of Paleontology , 64. 803-820.

— and pryor, w. a. 1984. Ophiomorpha: its morphologic, taxonomic and environmental significance.
Palaeogeography , Palaeoclimatology , Palaeoecologv, 23, 199-229.

and pemberton, s. c. 1985. Biogenic structures in outcrops and cores. I. Approaches to ichnology.
Bulletin of the Canadian Society of Petroleum Geologists, 33, 72-115.

— and fagerstrom, j. a. 1984. Morphological, ethological and environmental significance of the
ichnogenera Scoyenia and Ancorichnus. Journal of Paleontology. 58. 511-518.

fursich, f. t. 1974. Corallian (Upper Jurassic) trace fossils from England and Normandy. Stuttgarter Beitrdge
zur Naturkunde, Serie B ( Geologie and Palaontologie), 13, 1-52.

— 1981 Invertebrate trace fossils from the Upper Jurassic of Portugal. Communicagoes dos Geologicos de
Portugal, 67, 153-168.

— and heinberg, c. 1983. Sedimentology, biostratinomy, and palaeoecology of an Upper Jurassic offshore
sand bar complex. Bulletin of the Geological Society of Denmark, 32, 67-95.

gaillard, c. 1980. Megagyrolithes ardescensis n. gen. n. sp. trace fossile nouvelle du Valanginian d’Ardeche
(France). Geobios. 13, 465-471.

germs, G. J. b. 1972. Trace fossils from the Nama Group, South-west Africa. Journal of Paleontology, 46,
864-870.

gevers, j. w., brakes, l. a., Edwards, l. n. and marzolf, j. e. 1971. Trace fossils in the Lower Beacon
Sediments (Devonian) Darwin Mountains, Southern Victoria Land, Antarctica. Journal of Paleontology, 45,
81-94.

— and twomey, a. 1982. Trace fossils and their environment in Devonian (Silurian?) Lower Beacon strata
in the Asgard Range, Victoria Land, Antarctica. 639-649. In craddock, c. (ed.). Antarctic geoscience.
International Union of Geological Sciences, University of Wisconsin Press, Madison, 1172 pp.

gierlowski-kordesch, e. 1991. Ichnology of an ephemeral lacustrine/alluvial plain system: Jurassic East
Berlin Formation, Hartford Basin, USA. Ichnos, 1, 221-232.
glocker, e. f. 1841. Uber eine neue ratselhafte Versteinerung aus dem thonigen Spharosiderit der
Karpathensandsteinformation im Gebiete der Beskiden, nebst vorangeschickten Bemerkungen fiber die
Versteiner dieses Gebietes fiberhaupt. Nova Acta Academiae Caesareae Leopoldino-Carolinae Germanicae
Naturae Curiosorum, 19, 675-700.

goppert, h. r. 1841. Uber die fossile Flora der Quadersandsteinformation in Schlesien als erster Beitrag zur
Flora der Tertiargebilde. Nova Acta Academiae Caesareae Leopoldino-Carolinae Germanicae Naturae
Curiosorum, 19, 97-134.

Gordon, e. a. 1 988. Body and trace fossils from the middle-upper Devonian Catskill Magnafacies, southeastern
New York, U.S.A. 139-155. In McMillan, n. j., embry, a. f. and glass, d. j. (eds). Devonian of the World.
Vol. II. Proceedings of the 2nd International Symposium on the Devonian System, Calgary. Canadian
Society of Petroleum Geologists, 679 pp.

graham, j. r. and pollard, J. E. 1982. Occurrence of the trace fossil Beaconites antarcticus in the Lower
Carboniferous fluviatile rocks of County Mayo, Ireland. Palaeogeography, Palaeoclimatology, Palaeo-
ecology, 38, 257-268.

gumbel, c. w. 1879. Geognostiche Beschriebung des Fichtelgebirges mit dem Frankenwalde und dem westlichen
Vorlande 3. Perthes, Gotha, 698 pp.

hantzschel, w. 1962. Trace fossils and problematica. 178-245. In moorf,, r. c. (ed.). Treatise on invertebrate
paleontology. Part W. Miscellanea. Geological Society of America and University of Kansas Press, Boulder,
Colorado and Lawrence, Kansas, 259 pp.

KEIGHLEY AND PICKERILL: BEACON1TES

333

— 1972. Lebensspuren in den Kulm-Tonschiefern von Neustadt a.d. Weinstrasse. Oberrheinische geologische
Abhandlungen, 21, 107-115.

— 1975. Trace fossils and problematica. 1-269. In teichert, c. (ed. ). Treatise on invertebrate paleontology ,
Part W. Miscellanea . Supplement I . Geological Society of America and University of Kansas Press, Boulder,
Colorado and Lawrence, Kansas, 269 pp.

hakes, w. g. 1976. Trace fossils and depositional environments of four clastic units. Upper Pennsylvanian
megacyclothems, northeast Kansas. University of Kansas Paleontological Contributions , 63, 1-46.

— 1985. Trace fossils from brackish-marine shales. Upper Pennsylvanian of Kansas, U.S.A. 21-35. In
CURRAN, H. a. (ed.). Biogenic structures: their use in interpreting depositional environments. Special
Publication . Society of Economic Paleontologists and Mineralogists , 35, 1 347.

hall, J. 1852. Palaeontology of New York . Volume 2. State of New York, Albany, 362 pp.
hallam, a. 1960. Kulindridmus langi , a new trace fossil from the Lias. Palaeontology , 3, 64-68.
haskell, t. c., kennett, j. p. and prebble, w. m. 1965. Geology of the Brown Hills and Darwin Mountains,
Southern Victoria Land, Antarctica. Transactions of the Royal Society of New Zealand . 2, 231-248.
heer, o. 1865. Die Urwelt der Schweiz. F. Schulthess, Zurich, 622 pp.

1877. Flora fossilis Helvetiae. Die vonveltliche Flora der Schweiz. J. Wtirster and Co., 182 pp.

1881. Contributions a la flore fossile due Portugal. (Section des Travaux geologicptes du Portugal). Lisbon,
51 pp.

heinberg, c. 1970. Some Jurassic trace fossils from Jameson Land (East Greenland). 227-234. In crimes,
t. p. and harper, j. c. (eds). Trace fossils. Geological Journal Special Issue , 3, 1-347.

1974. A dynamic model for a meniscus filled tunnel (Ancorichnus n. ichnogen.) from the Jurassic Pecten
Sandstone of Milne Land, East Greenland. Rapport Gronlands Geologiske Undersogelse , 62, 1-20.

— and birkelund, t. 1984. Trace fossil assemblages and basin evolution of the Vardekloft Formation
(Middle Jurassic, central East Greenland). Journal of Paleontology . 58, 362-387.

horne, r. r. and Gardiner, p. r. R. 1973. A new trace fossil from non-marine Upper Palaeozoic red beds in
County Wexford and County Kerry, Ireland. Geologie en Mijnbouw, 52, 125-131
hosius, a. and von der marck, w. 1880. Die Flora der Westfalischen Kreideformation. Palaeontographica . 26,
125-236.

Howard, j. d. 1966. Characteristic trace fossils in Upper Cretaceous sandstones of the Book Cliffs and
Wasatch Plateau. Bulletin . Utah Geological and Mineralogical Survey. 80, 35-53.

— 1971. Comparison of the beach to offshore sequence in modern and ancient sediments. 149-183. In
Howard, j. D., valentine, j. w. and warme, J. e. (eds). Recent advances in paleoecology and ichnology. Short
course. American Geological Institute, 268 pp.

— 1972. Trace fossils as criteria for recognising shorelines in the stratigraphical record. 215-225. In rigby,
j. K. and hamblin, w. K. (eds). Recognition of ancient sedimentary environments. Special Publication .
Society of Economic Paleontologists and Mineralogists , 16, 1-340.

— and frey, r. w. 1984. Characteristic trace fossils in nearshore to offshore sequences. Upper Cretaceous
of east-central Utah. Canadian Journal of Earth Sciences. 21, 200-219.

Keighley, d. g. and pickerill, r. k. 1993. Nonmarine invertebrate ichnocoenoses from the Carboniferous of
Cape Breton Island. (Abstract). Atlantic Geology , 29, 83.
kern, j. p. 1978. Trails from the Vienna woods: paleoenvironments and trace fossils of Cretaceous to Eocene
flysch, Vienna, Austria. Palaeogeography , Palcieoclimatology. Palaeoecology. 23, 231-262.
ksiazkiewicz, m. 1977. Trace fossils in the flysch of the Polish Carpathians. Palaeontologia Polonica. 36, 1-208.
lessertisseur, j. 1955. Traces fossiles d’activite animale et leur signification paleobiologique. Memoires de la
Societe Geologique de France. 74, 1-48.

leszczynski, s. 1991. Trace-fossil tiering in flysch sediments: examples from the Guipuzcoan flysch
(Cretaceous-Paleogene), northern Spain. Palaeogeography. Palcieoclimatology . Palaeoecology. 88, 167-184.
liburnau, j. r. r. l. von 1990. Zur deutung der fossilen fucoiden-gattungen Taenidium und Gyrophyllites.

Denkschriften der Akademie der Wissenshaften Wien. 10. 523-583.
lockley, m. g., rindsberg, a. k. and zeiler, R. m. 1987. The paleoenvironmental significance of the nearshore
Curvolithus ichnofacies. Palaios. 2, 255-262.

Macdonald, d. i. m. 1982. Palaeontology and ichnology of the Cumberland Bay Formation, South Georgia.
Bulletin. British Antarctic Survey. 57, 1-14.

macleay, w. s. 1839. Note on the annelida. 699-701. In Murchison, r. i. The Silurian System. J. Murray,
London, 768 pp.

maillard, g. a. 1887. Considerations sur les fossiles decrits comme Algues. Memoires de la Societe
Paleontologique Suisse. 14, 1-40.

334

PALAEONTOLOGY, VOLUME 37

maples, c. G. and archer, a. w. 1989. The potential of Paleozoic nonmarine trace fossils for paleoecological
interpretations. Palaeogeography, Palaeoclimatology, Palaeoecology , 73, 185-195.

— and suttner, l. j. 1990. Trace fossils and marine-nonmarine cyclicity in the Fountain Formation
(Pennsylvanian: Morrowan/Atokan) near Manitou Springs, Colorado. Journal of Paleontology , 64,
859-880.

martino, r. l. 1989. Trace fossils from marginal marine facies of the Kanawha Formation (Middle
Pennsylvanian), West Virginia. Journal of Paleontology, 63, 389-403.
massalongo, a. 1855. Zoophycos , novum genus plant orum fossilium. Antonelli, Verona, 52 pp.
maulik, p. k. and chaudhuri, a. k 1983. Trace fossils from continental Triassic red beds of the Gondwana
sequence, Pranhita-Godavari valley. South India. Palaeogeography , Palaeoclimatology , Palaeoecology , 41,

1 7-34.

McCann, t. and pickerill, r. K. 1988. Flysch trace fossils from the Cretaceous Kodiak Formation of Alaska.
Journal of Paleontology, 62, 330-348.

McCarthy, b. 1979. Trace fossils from a Permian shoreface-foreshore environment, eastern Australia. Journal
of Paleontology, 53, 345-366.

McKELVEY, b. c., webb, p. n. and kohn, b. p. 1977. Stratigraphy of the Taylor and Lower Victoria Groups
(Beacon Supergroup) between the MacKay glacier and Boomerang ridge, Antarctica. New Zealand Journal
of Geology and Geophysics, 20. 813-863.

mikulas, r. 1990. Trace fossils from the Zahorany Formation (Upper Ordovician, Bohemia). Acta
Universitatis Carolinae, Geologica, 3, 307-335.

1992. Trace fossils from Early Silurian graptolitic shales of the Prague Basin (Czechoslovakia). Casopis
pro Mineralogii a Geologii, 37, 219-228.

1993. Trace fossils and ichnofacies of the Ordovician of the Prague Basin (central Bohemia, Czech
Republic). Boletin de la Real Sociedad Espahola de Historia Natural. Seccion Geologica , 88. 99-1 12.

miller, w. 1986. Discovery of trace fossils in Franciscan turbidites. Geology, 14. 343-345.

1989. Paleontology of Franciscan Flysch at Point Saint George. Northern California. 47-52. In aalto,
k. R. and harper, g. d. (eds). Geologic evolution of the northernmost Coast Ranges and western Klamath
Mountains, California. 28th International Geological Congress, Field Trip Guidebook T308, 82 pp.

— 1991a. Intrastratal trace fossil zonation, Cretaceous flysch of northern California. Ichnos. 1, 161-171.

— 1991 A Bathysiphonid (Protista: Foraminiferida) localities in Franciscan flysch. Northern California,
with a redescription of Bathysiphon aaltoi Miller, 1986. Tulane Studies in Geology and Paleontology, 24,
71-78.

muller, a. H. 1966. Neue Lebensspuren (Vestigia invertebratorum) aus dem Karbon und der Tnasa
Mitteldeutschlands. Geologie , 15, 712-725.

narbonne, G. m. 1984. Trace fossils in Upper Silurian tidal flat to basin slope carbonates of Arctic Canada.
Journal of Paleontology, 58, 398 415.

— and hofmann, h. j. 1987. Ediacaran biota of the Wernecke Mountains, Yukon, Canada. Palaeontology
30, 647-676.

— myrow, p. m., landing, e. and anderson, m. m. 1987. A candidate stratotype for the Precambrian-
Cambrian boundary. Fortune Head, Burin Peninsula, southeastern Newfoundland. Canadian Journal of
Earth Sciences, 24, 1277-1293.

nathorst, a. g. 1880. Om spar af nagra evertebrerade djur M.M. och deras paleontologiska betydelse.

Kongliga Svenska Vetenskaps- Akademiens Handlinger , 18, 1-104.
nilsen, t. h. 1982. Alluvial fan deposits. 49-86. In scholle, p. a. and spearing, d. (eds). Sandstone
depositional environments. Memoirs of the American Association of Petroleum Geologists, 31, 1-410.
ooster, w. a. 1869. Die organischen Reste der Zoophycos-SchChXen der Schweizer-Alpen. 3-35. In ooster.

w. a. and fischer-ooster, c. (eds). Protozoe Helvetica. H. Georg, Basel and Zurich, 43 pp.
orbigny, a. d’ 1842. Voyage dans I'Amerique meridionale. 3 (4) (Paleontologie). Levrault, Strasbourg. 188 pp.
o’sullivan, m. cooper, m. a., maccarthy, i. a. j. and forbes, w. H. 1986. The palaeoenvironment and
deformation of a ’ Beaconites ’ colony in the Old Red Sandstone at Gortnabrinna, southwest Ireland. Journal
of the Geological Society, London, 143, 897-906.

papp, a. 1941. Quergegliederte Rohrren aus dem Ober-Kriede-Flysch der Alpen. Palaeobiologica. 7, 314-317.
pearson, p. n. 1992. Walking traces of the giant myriapod Arthropleura from the Strathclyde Group (Lower
Carboniferous) of Fife. Scottish Journal of Geology, 28, 127-133.
pedersen, G. k. and surlyr, f. 1983. The Fur Formation, a late Paleocene ash-bearing diatomite from
northern Denmark. Bulletin of the Geological Society of Denmark, 32, 43-65.

KEIGHLEY AND PICKERILL: BEACONITES

335

Pemberton, s. G. and frey, r. w. 1982. Trace fossil nomenclature and the Planolites- Palaeophycus dilemma.
Journal of Paleontology , 56, 843-881.

— 1984. Ichnology of storm-influenced shallow marine sequence: Cardium Formation (Upper
Cretaceous) at Seebe, Alberta. 281-304. In stott, d. f. and glass, d. j. (eds). The Mesozoic of Middle North
America. Memoirs of the Canadian Society of Petroleum Geologists, 9, 1-573.

-and bromley, r. G. 1988. The ichnotaxonomy of Conostichus and other plug-shaped ichnofossils.
Canadian Journal of Earth Sciences , 25, 866-892.

perkins, b. f. and stewart, c. l. 1971 . Preface to stops 7 through 12. 56-70. In perkins, b. f. (ed. ). A field guide
to selected localities in Pennsylvanian, Permian, Cretaceous , and Tertiary rocks of Texas and related papers.
Society of Economic Paleontologists and Mineralogists Field Trip Guidebook April 1971, 148 pp.
pfeiffer, H. 1966. Die spurenfossilien des Kulms (Dinanats) und Devons der Frankenwalder Querzone
(Thiiringen). Jahrbuch fur Geologie , 2, 651-717.

pickerill, r. k. 1980. Phanerozoic flysch trace fossil diversity - observations based on an Ordovician flysch
ichnofauna from the Aroostook-Matapedia Carbonate Belt of northern New Brunswick. Canadian Journal
of Earth Sciences , 17, 1259-1270.

— 1989. Compaginatichnus : a new ichnogenus from Ordovician flysch of eastern Canada. Journal of
Paleontology , 63, 913-919.

— 1994. Nomenclature and taxonomy of invertebrate trace fossils. 3-42. In donovan, s. k. (ed.). The
palaeobiology of trace fossils. John Wiley and Sons, Chichester, 306 pp.

— donovan, s. k., doyle, E. n. and dixon, h. l. 1993. Ichnology of the Palaeogene Richmond Formation
of eastern Jamaica - the final chapter? Atlantic Geology, 29. 61-67.

— fyffe, L. R. and forbes, w. h. 1987. Late Ordovician-Early Silurian trace fossils from the Matapedia
Group, Tobique River, western New Brunswick, Canada. Maritime Sediments and Atlantic Geology , 23,
77-88.

— and harland, t. l 1988. Trace fossils from Silurian slope deposits. North Greenland. Rapport Gronlands
Geologiske Under sogelse, 137, 119-133.

-romano, m. and melendez, b. 1984. Arenig trace fossils from the Salamanca area, western Spain.
Geological Journal, 19, 249-269.

pienkowski, g. 1985. Early Liassic trace fossil assemblages from the Holy Cross Mountains, Poland: their
distribution in continental and marginal marine environments. 37—5 1 In curran, h. a. (ed.). Biogenic
structures: their use in interpreting depositional environments. Special Publication , Society of Economic
Paleontologists and Mineralogists, 35, 1-347.

plume, r. w. 1983. Environmental interpretation of the New Mountain Subgroup - marine or non-marine
(Beacon Supergroup). (Abstract). 233. In Oliver, r. l., james, p. r. and jago, j. b. (eds). Antarctic Earth
science. Cambridge University Press, Cambridge, 697 pp.
pollard, j. e. 1976. A problematical trace fossil from the Tor Bay Breccias of South Devon. Proceedings of
the Geologists' Association, 87. 105-108.

1988. Trace fossils in coal-bearing sequences. Journal of the Geological Society, London, 145, 339-350.

— steel, r. j. and undersund, e. 1982. Facies sequences and trace fossils in lacustrine/fan delta deposits,
Hornelen Basin, (Middle Devonian), western Norway. Sedimentary Geology 32, 63-87.

powichrowski, l. k. 1989. Trace fossils from the Helminthoid Flysch (Upper Cretaceous-Paleocene) of the
Ligurian Alps (Italy): development of deep marine ichnoassociations in fan and basin plain environments.
Eclogae Geologicae Helvetiae , 82, 385-411.

quatrefages, m. a. de 1849. Note sur la Scolicia prisca (A. de Q.), annelide fossile de la craie. Annates des
Sciences Naturelles , Serie 3, 12, 265-266.

ridgway, j. m. 1974. A problematical trace fossil from the New Red Sandstone of South Devon. Proceedings
of the Geologists' Association, 85, 511-517.

rindsberg, a. k. 1990. Commentary: ichnological consequences of the 1985 International Code of Zoological
Nomenclature. Ichnos, 1, 59-63.

romano, m. 1991. Lower to Middle Ordovician trace fossils from the Central Iberian Zone of Portugal and
Spain. 191-204. In barnes, c. r. and williams, s. h. (eds). Advances in Ordovician geology. Papers,
Geological Survey of Canada, 90-9, 326 pp.

SACCO, F. 1888. Note di paleoicnologia Italiana. Afti della Societd Italiana di Science Naturali , 31, 151-191.
sarkar, s. and chaudhuri, a. k. 1992. Trace fossils in Middle to Late Triassic fluvial redbeds,
Pranhita-Godavari Valley, South India. Ichnos, 2, 7-19.
schafer, w. 1956. Wirkungen der Benthos-Organismen auf der jungen Schichtverband. Senckenbergiana
Lethaea, 37, 183-263.

336

PALAEONTOLOGY, VOLUME 37

schafhautl, k. e. 1851. Geogiwstische Untersuchung des sudbayerischen Alpengebirges. Literarischen-artische
Anstalt, Munich, 208 pp.

scasso, R. a., olivero, e. b. and buatois, L. a. 1991. Lithofacies, biofacies and ichnoassemblage evolution of
a shallow submarine volcaniclastic fan-shelf depositional system (Upper Cretaceous), James Ross Island,
Antarctica. Journal of South American Earth Sciences, 4, 239-260.

SCHIMPER, w. p. 1869. Trade de Paleontologie vegetale. J. B. Bailliere et fils, Paris, 740 pp.

— and schenk, a. 1890. Palaeophytologie. 44-68, 233-234. In zittel, k. a. (ed). Handbuch Palaontologie.
Oldenbourg, Munich and Leipzig. 958 pp.

schroter, 1894. Notiz liber ein Taenidium aus dem Flysch von Ganey bei Seewis. Jahresbericht der
Naturforschenden Gesellschaft Graubiindens Chur , 37, 79-87.
scott, a. c., edwards d, and rolfe, w. d. I. 1976. Fossiliferous Lower Old Red Sandstone near Cardross,
Dumbartonshire. Proceedings of the Geological Society of Glasgow, 117, 4-5.
seilacher, a. 1955. Spuren und Fazies 1m Unterkambrium. 373-399. In schindewolf, o. h. and seilacher, a.
(eds). Beitrage zur Kenntnis des Kambriums in der Salt Range (Pakistan). Abhandlung der Mathematisch-
naturwissenschaftlichen Klasse. Akademie der Wissenschaften und der Literatur, Mainz, 10, 1-796.

1958. Flysch und Molasse als Faziestypen. Ec/ogae Geologicae Helvetiae, 51, 1062-1078.

1962. Paleontological studies on turbidite sedimentation and erosion. Journal of Geology, 70, 227-234.
1964. Biogenic sedimentary structures. 296-316. In imbrie, j. and Newell, n. (eds). Approaches to
paleoecology. Wiley, New York, 432 pp.

1978. Use of trace-fossil assemblages for recognizing depositional environments. 167-181. In basan,
p. b. (ed.). Trace fossil concepts. Short Course No. 5. Society of Economic Paleontologists and Mineralogists,

181 pp.

smith, r. m. h., mason, t. r. and ward, j. d. 1993. Flash-flood sediments and ichnofacies of the Late
Pleistocene Horneb Silts, Kuiseb River, Namibia. Sedimentary Geology, 85, 579-599.
squinabol, s. 1887. Contribuzioni alia flora fossile dei terreni Terziarii della Liguria. Bollettino della Societd
Geologica Italiana, 6, 545-561.

1891. Contribuzioni alia flora fossile dei terreni Terziarii della Liguria. I. Algae. Genova. 25 pp.
squires, r. L. and advocate, d. m. 1984. Meniscate burrows from Miocene lacustrine-fluvial deposits,
Diligencia Formation, Orocopia Mountains, southern California. Journal of Paleontology, 58, 593-597.
stanistreet, i. G., le blanc smith, G. and cadle, a. b. 1980. Trace fossils as sedimentological and
palaeoenvironmental indices in the Ecca Group (Lower Permian) of the Transvaal. Transactions of the
Geological Society of South Africa , 83, 333-344.

Stanley, d. c. a. and pickerill, r. k. 1993. Fustiglyphus annulatus from the Ordovician of Ontario, Canada,
with a systematic review of the ichnogenera Fustiglyphus Vialov 1971 and Rhabdoglyphus Vassoievich 1951.
Ichnos , 3, 57-67.

Stanley, k. o., chamberlain, c. k. and stewart, j. h. 1977. Depositional setting of some eugeosynclinal
Ordovician rocks and structurally interleaved Devonian rocks in the Cordilleran mobile belt, Nevada.
259-274. In stewart, j. h., Stevens, c. h. and fritsche, a. e. (eds). Paleozoic paleogeography of the western
United States. Pacific Coast Paleogeography Symposium, I. April 1977 . Society of Economic Paleontologists
and Mineralogists, 502 pp.

— and fagerstrom, j. a. 1974. Miocene invertebrate trace fossils from a braided river environment. Western
Nebraska, U.S.A. Palaeogeography, Palaeoclimatology, Palaeoecology, 15, 63-82.

Sternberg, K. M. von 1833. Versuch einer geognostisch-botanischen Darstellung der Flora der Vorwelt, 5-6.
Fleischer Fr., Leipzig, Prague, 80 pp.

tegan, j. r. and curran, h. a. 1992. Response of infaunal organisms represented by trace fossils to sea-level
changes in the Ordovician Black River and Trenton Group limestones, upstate New York (Abstract). A 154.
Geological Society of America Abstracts and Program, 1992 Annual Meeting. Cincinnati, 365 pp.
tevesz, m. j. s. and McCall, p. l. 1982. Geological significance of aquatic nonmarine trace fossils. 257-285. In
McCall, p. l. and tevesz, m. j. s. (eds). Animal-sediment relations. Plenum, New York, 336 pp.
thoms, r. E. and berg, t. h. 1985. Interpretation of bivalve trace fossils in fluvial beds of the basal Catskill
Formation (Late Devonian), Eastern USA. 13-20. In curran, h. a. (ed.). Biogenic structures: their use in
interpreting depositional environments. Special Publication, Society of Economic Paleontologists and
Mineralogists, 35, 1-347.

toots, h. 1967. Invertebrate burrows in the non-marine Miocene of Wyoming. Contributions to Geology
University of Wyoming, 6. 93-96.

trewin, N. H. 1976. Isopodichnus in a trace fossil assemblage from the Old Red Sandstone. Lethaia, 9, 29-37.

KEIGHLEY AND PICKERILL: BEACON ITES

337

valenzuela, m., garcia- ramos, j. c. and suarez de centi, c. 1986. The Jurassic sedimentation in Asturias (N.
Spain). Trabajos de Geologic/, 16. 121-132.

vassoievich, n. b. 1951. Usloviya obrazovaniya flisha. Gostoptekhizdat, Leningrad and Moscow, 240 pp.
vialov, o. s. 1962. Problematica of the Beacon Sandstone at Beacon Heights, West Antarctica. New Zealand
Journal of Geology and Geophysics, 5, 718-732.

— 1971. Redkie problematiki iz mesozoya Pamira i Kaukaza. Paleontologieheskii Sbornik, 1, 85-93.
walcott, c. D. 1890. The fauna of the Lower Cambrian or Olenellus Zone. Report of the United States
Geological Survey, 10, 515-760.

Walter, m. r., elphinstone, r. and heys, G. r. 1989. Proterozoic and Early Cambrian trace fossils from the
Amadeus and Georgina Basins, central Australia. Alcheringa, 13, 209-256.
webby, B. p. 1968. Devonian trace fossils from the Beacon Group of Antarctica. New Zealand Journal of
Geology and Geophysics, 11, 1001-1008.

wetzel, A. 1983. Biogenic structures in modern slope to deep-sea sediments in the Sulu Sea Basin (Philippines).

Palaeogeography, Palaeo climatology, Palaeoecology, 42, 285-304.
wheatcroft, r. 1986. Maestrichtian trace fossils from selected localities in the Chattachoochee River valley.
57-70. In reinhardt, J. (ed. ). Stratigraphy and sediment ology of continental, nearshore, and marine
Cretaceous sediments of the eastern Gulf coastal plain. SEPM/AAPG annual meeting 1986. Field Trip 3,

white, c. d 1929. Flora of the Hermit Shale, Grand Canyon, Arizona. Publications, Carnegie Institution of
Washington, 405, 1—221.

wiedman, l. a. and feldmann, r. m. 1988. Ichnofossils, tubiform body fossils, and depositional environment
of the La Meseta Formation (Eocene) of Antarctica. Memoirs, Geological Society of America , 169, 531-539.
wilckfns, o. 1947. Palaontologische und geologische Ergebnisse der Reise von Kohl-Larsen (1928-29) nach
Siid-Geogien. Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft, 474, 1-75.
williams, b. p. j. 1978. Locality B18 - Freshwater West, Dyfed. 95-98. In friend, p. f. and williams, b. p. j.
(eds). A field guide to selected outcrop areas of the Devonian of Scotland, the Welsh Borderland, and South
Wales. Palaeontological Association, London, 106 pp.

woolfe, k. j. 1990. Trace fossils as paleoenvironmental indicators in the Taylor Group (Devonian) of
Antarctica. Palaeogeography, Palaeoclimat ology, Palaeoecology, 80, 301-310.
zenker, j. c. 1836. Historisch-topographisches Taschenbuch von Jena und seiner Umgebung besonders in
naturwissenschaftlicher und medicinischer Beziehung. Wackenhoder, Jena, 338 pp.

99 pp.

Typescript received 29 June 1993
Revised typescript received 27 January 1994

D. G. KEIGHLEY
R. K. PICKERILL

Department of Geology
University of New Brunswick
Fredericton, Canada E3B 5A3

A LARGE OWL FROM THE PALAEOGENE OF

FRANCE

by CECILE MOURER-CHAUVIRE

Abstract. Strigiformes have a very long palaeontological history and were much more diversified in the past
than they are at present. Berruornis orbisantiqui gen. et sp. nov., from the Upper Palaeocene (Thanetian) of the
Reims area, is a large owl, with a stout tarsometatarsus. It belongs to the Sophiornithidae, previously described
from the Upper Eocene to Upper Oligocenc of the Phosphorites du Quercy, France. The comparison of the
distal part of the tarsometatarsus in different extinct forms shows a general trend towards a progressively more
semizygodactyl foot, with an internal trochlea more posteriorly oriented, and an external trochlea more
medially incurved; correspondingly the size of the internal trochlea compared with the middle trochlea
decreases.

Bird remains found in the Palaeocene deposits of the area of Reims, in north-eastern France, have
been known for over a century (Lemoine 1878-1881). They mainly include large forms belonging
to Gastomis and Remiornis. Part of this material was recently revised by Martin (1992), who placed
Gastornis in the order Gastornithiformes and created for Remiornis the new order Remiornithi-
formes, in the Palaeognathae.

Excavations in the region of Cernay-les-Reims, and mainly in Mont Berru, have been resumed
by D. E. Russell who has collected a large quantity of mammals and a few bird remains. The dating
of these localities has been carried out by Russell (1964) on the mammal faunas. Apart from the
large forms, the fossil avifauna includes Gruiformes (Cariamidae and Messelornithidae ; Mourer-
Chauvire in press), Charadriiformes, ‘Form-Family’ Graculavidae, and a large form of owl, which
is relatively abundant, and is the subject of the present study.

The osteological terminology generally follows Howard (1929) and, when necessary, Baumel
(1979u). Institutional abbreviations are: AMNH, American Museum of Natural History, New
York, USA; KUMNH, Kansas University Museum of Natural History, Lawrence, Kansas, USA;
USNM, National Museum of Natural History, Washington D.C., USA.

SYSTEMATIC PALAEONTOLOGY

Class aves Linnaeus, 1758
Order strigiformes Wagler, 1830
Family sophiornithidae Mourer-Chauvire, 1987

Type-genus. Sophiornis Mourer-Chauvire, 1987

Other genera. Berruornis gen. nov.

Distribution. Palaeocene of the Reims area, and Upper Eocene to Upper Oligocenc of Phosphorites du Quercy,
France.

Genus berruornis gen. nov.

Type-species. Berruornis orbisantiqui sp. nov.
| Palaeontology, Vol. 37, Part 2, 1994, 339-348, I pi. |

© The Palaeontological Association

340

PALAEONTOLOGY, VOLUME 37

Distribution. Upper Palaeocene of the Reims area, Thanetian, Mammal Reference-Level MP 6 (Schmidt-
Kittler 1987).

Diagnosis. Tarsometatarsus showing the characteristic features of the Sophiornithidae, but differing
from Sophiornis by 1, proximal part of internal calcaneal ridge situated slightly distally compared
with the internal cotyla; 2, internal calcaneal ridge perforated by a foramen. Features unknown in
Sophiornis are: 3, middle trochlea relatively narrow and slightly longer distally than internal
trochlea; 4, external trochlea much shorter distally than middle trochlea, very narrow and
prolonged by a posteriorly directed wing. On the distal part of the tibiotarsus; 5, internal condyle
relatively narrow and not flattened; 6, supracondylar fossa on the internal side of the shaft.

Derivation of name. " Berru', from the locality of Mont Berru, and " ornis', bird.

Berruornis orbisantiqui sp. nov.

Plate 1, figures 1-14

Holotype. Museum national d’Histoire naturelle de Paris, R 4155, incomplete right tarsometatarsus; from
Mont Berru, Reims area, France.

Material (all in the same collection as the holotype). Mont Berru : BR 11186, left tibiotarsus, fragment of distal
part; BR 11195, right tarsometatarsus, fragment of proximal part and shaft; BR 12482-12483, right
tarsometatarsus, fragments of proximal and distal parts; BR 14571, right tarsometatarsus, shaft and
incomplete distal part, subadult; Cernay area: L 3096, cast of a complete right tarsometatarsus.

Horizon and localities. Upper Palaeocene, Thanetian, Mammal Reference-Level MP 6 (Schmidt-Kittler 1987),
Reims area (Cernay and Mont Berru), Marne Departement, France.

Diagnosis. As for the genus.

Dimensions. See Tables 1-2.

Derivation of name. From "or bis \ the world, and ’ antiquus ’, old, because this genus (and family), are known
only from the Old World.

Description and comparisons

Tarsometatarsus (PI. 1, figs 1-13). The tarsometatarsus corresponds to a large form, approximately of the size
of the Recent Eagle Owl ( Bubo bubo). It is very stout, with a wide shaft. This feature is not so well developed
in BR 14571, which is subadult.

EXPLANATION OF PLATE I

Figs 1 14. Berruornis orbisantiqui , gen. nov. sp. nov.; collection of the Museum national d'Histoire naturelle
de Paris. 1 -5, R 4155, right tarsometatarsus holotype; 1, anterior view; x 1. 2, internal view showing the
break of the internal calcaneal ridge at the level of the foramen; x 1. 3, posterior view; x 1. 4, proximal view;
x L5. 5, distal view; x L5. 6-10, L 3096, cast of right tarsometatarsus; 6, anterior view; x 1. 7, internal view
showing the break of the internal calcaneal ridge at the level of the foramen; x 1. 8, posterior view; x 1.
9, proximal view; x 1-5. 10, distal view; x L5. 11-12, BR 12483; fragment of proximal part of right
tarsometatarsus; 11, internal view, showing the foramen in the internal calcaneal ridge; x 1-5. 12, medial
view of the internal calcaneal ridge; x 1-5. 13, BR 12482; fragment of distal part of right tarsometatarsus;
internal trochlea; internal view; x L5. 14, BR 11186; left tibiotarsus, incomplete distal part, anterior view;
x 1 .

Figs 15-16. Sophiornis quercynus , collection Musee Guimet d’Histoire naturelle de Lyon, PQ 1202. 15, anterior
view; x 1. 16, posterior view; x 1.

Fig. 17. Palaeoglaux perrierensis , collection Universite des Sciences et Techniques du Languedoc, Montpellier,
PRR 2576; right tarsometatarsus, paratype; distal view; x 2.

PLATE 1

MOURER-CHAUVIRE, Berruornis , Sophiornis , Palaeoglaux

342

PALAEONTOLOGY, VOLUME 37

Table 1. Dimensions of the tarsometatarsus in Sophiornithidae, in mm. (a) Proximal depth from anterior edge
of internal cotyle to posterior end of internal calcaneal ridge.

Tarsometatarsus

Berruornis orbisantiqui

Sophiornis
quercynus
PQ 1202

R4155

Holotype

L3096

BR 12482
and 12483

BR11195

BR14571

juv.

Total length

69-8 as pres.

68-5

—

—

—

74 0 as pres.

Prox. width

19-9

1 9-4 as pres.

—

—

—

19 9

Prox. depth (a)

13-7

1 5-0 as pres.

17-0

—

—

17-2

Width shaft in the middle

14-2 as pres.

16-2

—

1 1-8

9-1

12 1

Depth shaft in the middle

6-8 as pres.

6-2

—

7-2

5-2

6-0

Distal width

—

24 1

—

—

—

23-7

Distal depth

—

14 0

—

—

—

13-0 as pres.

Width int. trochlea

12-0

12-5

15-8

—

11-7

1 3-3

Width middle trochlea

—

7-5

—

—

4-3

—

Width ext. trochlea

—

4-8

— -

—

—

—

Depth int. trochlea

9-4

9-7

1 1-7

—

8-7 as pres.

8-7

Depth middle trochlea

—

10 0

—

—

7-0 as pres.

—

Depth ext. trochlea

—

12-7

—

—

—

—

Table 2. Dimensions of the distal part of tibiotarsus in some fossil owls, in mm. (a) In Rich (1982) the
measurements are given for Protostrix, but this generic name is a junior synonym of Minerva (Mourer-
Chauvire 1983). * measured from the illustrations in Fischer (1983).

Minerva

Minerva

Berruornis

leptosteus

antiqua

Oligostrix

orbisantiqui

(Rich 1982)

(Mourer-Chauvire

rupelensis

Tibiotarsus

BR 11186

(a)

1983)

(Fischer 1983)

Distal width (as preserved)

14-2

13-5

15-8-15-9

7-0

Distal depth (as preserved)

13-3

—

—

6-0

Width internal condyle

50

—

6-3— 6-8

3-3*

Width of shaft at the level of

8-6

—

—

3-7*

insertion of retinaculum

extensorium tibiotarsi

Depth of shaft at the same level

5-8

—

—

—

In proximal view, the proximal articular surface is anteroposteriorly narrow at the level of the intercotylar
prominence, and this intercotylar prominence is medio-laterally elongated. These features are clearly visible on
the holotype, and are slightly different on L 3096 which seems to have suffered some deformation. On the
anterior face, the proximal foramina are very small. The internal foramen is situated slightly more distally than
the external foramen. It is not possible to see the tubercle for m. tibialis anticus because of the poor preservation
of the anterior surface of the shafts.

The tarsometatarsus of Berruornis exhibits the characteristic features of the Sophiornithidae: on the anterior
face, a shallow depression below the proximal articular surface but no real anterior metatarsal groove; no
ossified supratendinal bridge; posterior metatarsal groove very shallow; cross-section of shaft rectangular;
trochleae arranged along a weakly curved line, but differs from the type genus Sophiornis by the following
features: proximal articular surface very narrow at the level of the intercotylar prominence (wider in
Sophiornis ); proximal surface of the internal calcaneal ridge situated distally compared with the internal cotyle
(at the same level as internal cotyle in Sophiornis ); internal calcaneal ridge wide, straight in internal view, and
pierced by a foramen (thin, semi-circular in shape in internal view, proximo-distally elongated, not perforated

MOURER-CHAUVIRE: PALAEOGENE OWL

343

in Sophiornis ); proximal surface of the external calcaneal ridge distinct from the external cotyle (as an extension
of the external cotyla in Sophiornis)', external calcaneal ridge relatively thin and posteriorly directed (wider and
postero-externally directed in Sophiornis) ; anterior infracotylar fossa slightly indicated, with traces of insertion
of a non-ossified supratendinal bridge (almost missing, with no visible traces of supratendinal bridge in
Sophiornis); posterior metatarsal groove shallow (almost completely missing in Sophiornis); internal trochlea
directed internally and posteriorly, with an angle of 61° compared with the anterior face (directed more
posteriorly, with an angle of 68° in Sophiornis) (Text-fig. I); deep fossa on the internal face of the internal

text-hg. 1. Tarsometatarsus in distal view. Angle of the internal and external trochleae compared with a
straight line drawn through the anterior part of these trochleae. a, Sophiornis quercynus , PQ 1202, holotype,
left tarsometatarsus. Hatched area represents missing parts, b, Berruornis orbisantiqui , L 3096; right

tarsometatarsus.

trochlea (shallower fossa in Sophiornis); articular surface of the internal trochlea forming a spike above the
wing of the trochlea, this spike also exists in the protostrigids Eostrix martinellii (Martin and Black, 1972) and
Minerva leptosteus (Rich, 1982) (spike much less developed in Sophiornis); external trochlea directed almost
posteriorly, with an angle of 76° compared with the anterior face (not so posteriorly directed, with an angle
of 65° in Sophiornis) (Text-fig. 1); in posterior view, proximal part of external trochlea situated more
proximally than the proximal part of internal trochlea (proximal part of external and internal trochleae
situated at the same level in Sophiornis); metatarsal facet for digit 1 well marked, in particular on BR 1 1 195,
and situated on the internal side of the shaft (more weakly indicated and situated on the postero-mternal angle
of the shaft in Sophiornis).

In Berruornis, on the posterior face, the external proximal foramen is situated at the base of the external
calcaneal ridge and the internal proximal foramen on the medial side of the internal calcaneal ridge, but the
latter is prolonged by a foramen which goes through the calcaneal ridge and which is clearly visible on fragment
BR 12483 (PI. 1, figs 1 1-12). The presence of this foramen has produced a weakness in the internal calcaneal
ridge, which is broken at the level of this foramen on the other two specimens (R 4155 and L 3096) (PI. 1, figs
2, 7). In the Recent Strigiformes as well as in Sophiornis, the arteriola tarsalis plantaris (Baumel 19796, p. 372,
annt. 79) of the inner side goes directly through the internal proximal foramen from the anterior face to the
internal side of the internal calcaneal ridge, while in Berruornis this arteriola first goes out on the medial side
of the internal calcaneal ridge, then crosses through the calcaneal ridge.

At the distal end, in Berruornis as in Sophiornis, the internal trochlea is very wide and strong. The external
and middle trochleae are incompletely preserved in Sophiornis and are only known in Berruornis. The internal
trochlea is slightly shorter than the middle trochlea, while the external trochlea is considerably shorter. In distal
view (Text-fig. 2b), the middle trochlea is slightly asymmetrical, with an external border slightly deeper than
the internal border. The external trochlea is narrow, with a weakly developed wing, and is directed almost
posteriorly.

On the grounds of the morphological differences between the form from Mont Berru and that from Quercy,
it seems justified to place them in two different genera. Moreover they are separated by a long interval of lime.
The Mont Berru form is dated from the Thanetian (between 58 and 54 Ma; Savage and Russell 1983). The age
of the Quercy form is not known accurately but lies between the beginning of the Upper Eocene (about 38 Ma),

344 PALAEONTOLOGY, VOLUME 37

text-fig. 2. The right tarsometatarsus, in distal view, in different species of fossil and Recent Strigiformes. All
the figures have been brought to the same distal width, and the figures concerning left tarsometatarsi have been
reversed in order to make the comparison easier. This figure shows the evolution of the external trochlea which
becomes increasingly posteriorly elongated and internally curved in the course of time. The internal trochlea,
strongly developed in Sophiornithidae ( B) and Protostrigidae (d-e) becomes proportionally smaller and its
orientation, internal in Sophiornithidae, becomes increasingly posterior, a, Ogygoptynx wetmorei , AMNH
2653; right tarsometatarsus; Palaeocene; x6-5. after Rich and Bohaska 1976. b, Berruornis orbisantiqui gen.
et sp. nov. L 3096; right tarsometatarsus; Palaeocene: x2-3. c. Palaeoglaux perrierensis , PRR 2576; right
tarsometatarsus; Eocene; x4-9. d, Eostrix martinellii, KUMNH 16601; left tarsometatarsus (reversed);
Eocene: x 6. after Martin and Black 1972. e, Minerva leptosteus , AMNE1 2629, right tarsometatarsus; Eocene;
x4. after Rich 1982. f, Necrobyas harpax , QU 16298; left tarsometatarsus (reversed); Oligocene; x 5-66.
G, Strix aluco, Lyon 252-1; right tarsometatarsus; Recent; x4-8. h, Tyto alba , Lyon 245-1; right tarso-
metatarsus; Recent; x 5 8. i, Phodilus badius, USNM 20310; left tarsometatarsus (reversed); Recent; x4-66.

and the end of the Upper Oligocene (about 24 Ma; ages after Harland et a!. 1989). The interval of time which
separates the two forms therefore is from a minimum of 16 Ma to a maximum of 34 Ma.

Tibiotarsus (PI. 1. fig. 14). On the distal part of tibiotarsus BR I 1 186, only part of one condyle is preserved
(only its anterior half). The presence of the tubercle for the attachment of retinaculum extensorium tibiotarsi
(Baurnel 1979«), just below the break of the bone, and the presence on the external side, above the missing
condyle, of a longitudinal ridge which limits backwards the groove for peroneus profundus, make it possible to
state that this tibiotarsus is from the left side, and that the preserved condyle is the internal one.

MOURER-CHAUVIRE: PALAEOCENE OWL

345

This condyle is relatively narrow and anteriorly and distally rounded. It is very different from the condition
found in the Protostrigidae, where the internal condyle is strongly widened and flattened. This can also be
confirmed by the shape of the internal cotyle of the tarsometatarsus, which is hollow. This internal cotyle could
not have corresponded to a flattened internal condyle of tibiotarsus.

On the anterior face of the shaft, above the intercondylar groove, there is a shallow, but well indicated,
supracondylar fossa, situated on the internal side of the shaft. Unlike the condition of the internal condyle, the
presence of a supracondylar fossa situated on the internal side of the shaft is known in all the members of the
family Protostrigidae (Wetmore 1933, 1937, 1938; Fischer 1983; Mourer-Chauvire 1983). The position and
development of the supracondylar fossa is different in the other Recent and fossil Strigiformes (Mourer-
Chauvire 1987). The intercondylar groove is relatively wide, wider than in Recent Strigidae and Tytomdae.
On the posterior face, the condyles are not preserved, the posterior intercondylar groove is wide and there is
no supracondylar fossa, while this fossa exists in the Recent Strigiformes.

Comparison with other fossil Strigiformes

(a) Bradycnemidae. The oldest fossils described as Strigiformes form the basis of the extinct family
Bradycnemidae (Harrison and Walker 1975) from the Upper Cretaceous of Romania, but
subsequent authors (Brodkorb 1978; Olson 1985) have argued that the remains from which this
family was created are not avian.

(b) Ogygoptyngidae . The oldest strigiform presently known is Ogygoptynx (Ogygoptyngidae), from
the Palaeocene (TifTanian) of Colorado (Rich and Bohaska 1976, 1981). Several hypotheses have
been proposed for the correlations between the Palaeocene and Lower Eocene of Europe and North
America in recent years (Savage and Russell 1983; Russell et a/. 1990), but according to the latter
(1990, p. 29, my translation): ‘it seems that there are some indications that a large part of the Middle
and Late Paleocene (Torrejonian, TifTanian, Clarkforkian) of North America could be equivalent
to the Thanetian of Europe’. In this case, Berruornis , from the Thanetian, could compete with
Ogygoptynx as the world’s oldest owl.

The Ogygoptyngidae differ from the Sophiornithidae in the following main characteristics:
tarsometatarsus slender and elongated; deep anterior metatarsal groove; proximal articular surface
in proximal view shaped like a parallelogram; internal trochlea decidedly longer than middle
trochlea; in distal view, curvature across the trochleae much more developed (Text-fig. 2a); in distal
view, external trochlea not smoothly rounded but slightly grooved laterally (Rich and Bohaska
1976, 1981).

(c) Protostrigidae. This family is represented by three genera: Eostrix , from the Lower and Middle
Eocene of the United States, and to which has been attributed a pedal phalanx from the Lower
Eocene of England (Harrison 1980; Olson 1985), Minerva , from the Middle and Upper Eocene of
the United States (Mourer-Chauvire 1983; Olson 1985), and Oligostrix , from the Lower Oligocene
of Germany (Fischer 1983). It is the only extinct family of Strigiformes which is known both from
Europe and North America. These forms are mainly represented by distal parts of tibiotarsus, distal
parts of tarsometatarsus, and pedal phalanges.

The Protostrigidae differ from the Sophiornithidae because, in the former the internal condyle of
tibiotarsus is distinctly widened and flattened; on the tarsometatarsus the anterior metatarsal
groove is deep across the full width of the proximal end (Mourer-Chauvire 1983), and the curvature
across the trochleae is much more pronounced (Text-fig. 2d-e). The internal trochlea is strongly
developed but not to such an extent as in Sophiornithidae.

(d) Palaeoglaucidae. This family consists of one genus, Palaeog/aux, and two species, P. artophoron
(Peters 1992), from the Middle Eocene of Messel, in Germany, and P. perrierensis (Mourer-
Chauvire 1987), from the Upper Eocene of Phosphorites du Quercy. Palaeoglaux differs from the
Sophiornithidae by its tarsometatarsus which is more slender and has an anterior metatarsal groove
(Peters 1992). In P. perrierensis the distal part of the tarsometatarsus strongly widens at the level
of the internal and external trochleae, which distinctly project on the internal and external sides

346

PALAEONTOLOGY, VOLUME 37

respectively, and the internal trochlea is almost as wide as the middle trochlea, while in
B. orbisantiqui the internal trochlea is much wider (PI. 1, fig. 17; Text-fig. 2c).

(e) Other fossil forms from the Phosphorites du Quercy. The other extinct genera from the
Phosphorites du Quercy such as Necrobyas Milne-Edwards, 1892, Nocturnavis Mourer-Chauvire,
1987, Palaeobyas Mourer-Chauvire, 1987, Palaeotyto Mourer-Chauvire, 1987, and Selenornis
Mourer-Chauvire, 1987, have been described in detail and ascribed to the Recent family Tytonidae
(Mourer-Chauvire 1987).

(f) Eoglaucidium pallas Fischer , 1987. This genus and species, from the Middle Eocene of Geiseltal,
Mammalian Reference-levels MP 11, 12, and 13 (Schmidt-Kittler 1987) is known from eight humeri.
It was classified in the Recent family Strigidae but, according to Peters (1992) and to Mlikovsky
(1992), who is studying other elements associated with these humeri, it may belong to the
Coraciiformes.

(g) Genus incertae sedis Eupterornis Lemoine , 1878. Lemoine (1878-1881, 1884) described from the
locality of Cernay the genus Eupterornis , the type of which is a distal part of ulna, strongly flattened
and which looks somewhat like the ulna of a loon (Gaviidae). Its systematic position is not yet
defined (Lambrecht 1933; Brodkorb 1963; Olson 1985), and it is better to consider it as incertae
sedis. The illustrations of the holotype and of a wing phalanx referred to the same species, do not
show any resemblances to Strigiformes and therefore the tibiotarsus and tarsometatarsi from
Cernay and Mont Berru cannot be attributed to that genus.

(h) Strigiform from the Upper Palaeocene of Kazakhstan. A pedal phalanx of a large-sized strigiform
has been reported and illustrated by Nessov (1992, fig. 5j and k) from the Upper Palaeocene
(Landenian) of the Zhylga locality in Kazakhstan. From its size, this phalanx could correspond to
Berruornis. This indicates that large owls were also present during the Palaeocene in Central Asia.

CONCLUSIONS

The Strigiformes, which are represented today by only two families, were already very diverse
during the Palaeocene, where they are represented by two families, the Ogygoptyngidae, which are
known only from one small form, and the Sophiornithidae, which are, on the contrary, large forms.

This diversification continued during the Eocene, where four families are known, three extinct,
namely the Protostrigidae, the Sophiornithidae, and the Palaeoglaucidae, and the Recent family
Tytonidae, which is represented in the Phosphorites du Quercy deposits by two extinct subfamilies,
the Necrobyinae, and the Selenornithinae. The Protostrigidae are themselves very diverse and could
belong to two distinct families (Olson 1985).

These families, with the exception of Palaeoglaucidae, persisted during the Oligocene, and there
is in the United States a large quantity of Oligocene owl material which has not yet been studied
(Olson 1985). Among the Recent suprageneric taxa, the Tytoninae and the Strigidae appear in
Europe from the Lower Miocene (Mourer-Chauvire 1987). The Strigiformes have, therefore, a very
long palaeontological history and in the past they were more diverse than they are now.

The Recent Strigiformes have a semizygodactyl foot, which means that digit IV can be directed
forwards or backwards. According to Raikow (1985, p. 119): ‘Presumably this ability enhances the
functional potential of the foot’.

In the Ogygoptyngidae the shape of the foot, in distal view, is very peculiar (Text-fig. 2a). It seems
that digits II and IV could be strongly splayed and that the foot could have developed into an
ectropodactyl type, similar to that which can be observed in the Piciformes when climbing on a
vertical surface (Raikow 1985).

In families other than Ogygoptyngidae it is possible to see, from the Sophiornithidae (Text-fig.
2b), then the Protostrigidae (Text-fig. 2d and 2e), then the Palaeoglaucidae (Text-fig. 2c), to the

MOURER-CHAUVIRE: PALAEOCENE OWL

347

Necrobyinae (Text-fig. 2f) and the Recent forms, Strigidae (Text-fig. 2g), Tytoninae (Text-fig. 2h),
Phodilinae (Text-fig. 2i), an evolution towards a more and more semizygodactyl type, with the
trochlea for digit IV directed first extero-posteriorly, then more and more posteriorly, then more
and more curved medially, the maximum being reached in the Recent genus Phodilus (Text-fig. 2i).
At the same time the trochlea for digit II, strongly developed in the Sophiornithidae and
Protostrigidae, becomes proportionally smaller, and its orientation also changes. The wing of this
trochlea is predominantly oriented internally in the Sophiornithidae and becomes more posteriorly
oriented in the other families. This can be explained from the functional point of view either by
progressively more pronounced perching habits, or by a better adaptation to the capture of prey.
In the first case, that would indicate that the Sophiornithidae were more terrestrial than the Recent
Strigiformes.

Acknowledgements . I thank Donald E. Russell for allowing me to study this material which he collected.

REFERENCES

baumel, J. J. 1979a. Osteologia. 53-121. In baumel, j. j. (ed.). Nomina anatomica avium. An annotated
dictionary of birds. Academic Press, London and New York, 637 pp.

— 19796. Systema cardiovasculare. 343 — 407 . In baumel, j. j. (ed.). Nomina anatomica avium. An annotated
dictionary of birds. Academic Press, London and New York, 637 pp.

brodkorb, p. 1963. Catalogue of fossil birds. Part 1 (Archaeopterygiformes through Ardeiformes). Bulletin of
the Florida State Museum , 7, 179-293.

1978. Catalogue of fossil birds. Part 5 (Passeriformes). Bulletin of the Florida State Museum , 23, 139-228.
fisher, k. 1983. Oligostrix rupelensis n. gen., n. sp., eine neue Ureule (Protostrigidae, Strigiformes, Aves) aus
dem marinen Mitteloligozan des Weisselsterbeckens bei Leipzig (DDR). Zeitschrift fur Geologische
Wissenschaften , 11, 483-487.

- — 1987. Eulenreste ( Eoglaucidium pallas nov. gen., nov. sp., Strigiforme, Aves) aus der mitteleozanen
Braunkohle des Geiseltals bei Halle (DDR). Mitteilungen aus dem Zoologischen Museum Berlin 63 ,
Supplement : Annalen fur Ornithologie 11, 137-142.

harland, w. b., Armstrong, r. l., cox, a. v., craig, l. E., smith, a. G. and smith, d. G. 1989. A geologic time
scale 1989. Cambridge University Press, Cambridge, 263 pp.

Harrison, c. J. o. 1980. A small owl from the Lower Eocene of Britain. Tertiary Research , 3, 83-87.

— and walker, c. a. 1975. The Bradycnemidae, a new family of owls from the Upper Cretaceous of
Romania. Palae ontology. 18, 563-570.

Howard, H. 1929. The avifauna of Emeryville shellmound. University of California Publications , Zoology , 32,
301-394.

lambrecht, K. 1933. Handbuch der Palaeornithologie. Borntraeger, Berlin, 1024 pp.

lemoine, v. 1878-1881 . Recherches sur les oiseaux fossiles des terrains tertiaires inferieurs des environs de Reims.
Premiere partie, Imprimerie et Lithographic Keller, Reims, 1-69; Deuxieme partie, Imprimerie et
Lithographic Matot-Braine, Reims, 75-170.

— 1884. Presentation de pieces du squelette d'Eupterornis , de Gastornis et de Pleuraspidotherium. Bulletin
de la Societe Geologique de France , 3 serie , 12, 357.

Linnaeus, c. 1758. Systema naturae. (10th edn). L. Salvii, Stockholm, 823 pp.

martin, L. D. 1992. The status of the Late Paleocene birds Gastornis and Remiornis. Natural History Museum
of Los Angeles County , Sciences Series, 36, 97-108.

— and black, c. c. 1972. A new owl from the Eocene of Wyoming. Auk, 89, 887-888.
milne-edwards, a. 1892. Sur les oiseaux fossiles des depots eocenes de phosphate de Chaux du Sud de la

France. Comptes Rendus du second Congres ornithologique international, Budapest 1892, 60-80.
mlikovsky, j. 1992. The present status of knowledge of the Tertiary birds of Central Europe. Natural History
Museum of Los Angeles County , Sciences Series , 36, 433-458.
mourer-chauvire, c. 1983. Minerva antiqua (Aves: Strigiformes), an owl mistaken for an edentate mammal.
American Museum Novi fates , 2773, 1-11

— 1987. Les Strigiformes (Aves) des Phosphorites du Quercy (France): Systematique, Biostratigraphie et
Paleobiogeographie. Documents des Laboratoires de Geologie de Lyon, 99, 89-135.

— in press. The Messelornithidae (Aves: Gruiformes) from the Paleogene of France. Courier Lorschungs-
institut Senckenberg.

348

PALAEONTOLOGY, VOLUME 37

nessov, l. a. 1992. Mesozoic and Paleogene birds of the USSR and their paleoenvironments. Natural History
Museum of Los Angeles County , Sciences series, 36, 465^478.
olson, s. L. 1985. The fossil record of birds. Avian Biology, 8, 79-252.

peters, d. s. 1992. A new species of owl (Aves: Strigiformes) from the Middle Eocene Messel Oil Shale.

Natural History Museum of Los Angeles County, Sciences series , 36, 161-169.
raikow, r. j. 1985. Locomotor system. Form and function in birds. Academic Press, London and New York,
3, 57-147.

rich, p. v 1982. Tarsometatarsus of Protostrix from the mid-Eocene of Wyoming. Auk, 99. 576-579.

- and bohaska, d. j. 1976. The world’s oldest owl: a new strigiform from the Paleocene of southwestern
Colorado. Smithsonian Contributions to Paleobiology, 27, 87-93.

- 1981. The Ogygoptyngidae, a new family of owls from the Paleocene of North America. Alcheringa , 5,
95-102.

russell, d. E. 1964. Les mammiferes paleocenes d'Europe. Memoires du Museum national if Histoire naturelle
de Paris, serie C, 13, 1-324.

broin, f. de, galoyer, a., gaudant, j., GiNGERiCH. p. d. and rage, j. c. 1990. Les Vertebres du
Sparnacien de Meudon. Bulletin cf Information des Geologues du Bassin de Paris , 27, 21-31.
savage, d. E. and russell, d. E. 1983. Mammalian paleofaunas of the world. Addison-Wesley, London,
Amsterdam, Don Mills, Ontario, Sydney and Tokyo, 432 pp.

SCH midi- k itilir, n. 1987. European reference levels and correlation tables. Miinchner Geowissenschaftliche
Abhandlungen, Reihe A, Geologie und Paldontologie. 10, 13-19.
wagler, J. G. 1830. Naturliches System der Amphibien mit vorangehender Classification der Sdugthiere und
Vogel. Ein Beitrage zur vergleiclienden Zoologie. Munich, Stuttgart and Tubingen, vi + 354 pp.
wetmore, a. 1933. The status of Minerva antique/, Aquila ferox and Aquila lydekkeri as fossil birds. American
Museum Novitates, 680. 1 4.

- 1937. The systematic position of Bubo leptosteus Marsh. Condor , 39, 84-85.

- 1938. Another fossil owl from the Eocene of Wyoming. Proceedings of the United States Museum, 85
(3031), 27-29.

CECILE MOURER-CHAUVIRE

Centre de Paleontologie stratigraphique et Paleoecologie
associe au CNRS (URA 1 1)

Universite Claude Bernard - Lyon 1

Typescript received 28 July 1993 27-43 boulevard du 11 novembre 1918

Revised typescript received 22 October 1993 69622 Villeurbanne Cedex, France

RE-INTERPRETATION OF TEREBRATU LI DE
PHYLOGENY BASED ON IMMUNOLOGICAL DATA

by K. ENDO, G. B. CURRY, R. QUINN, M. J. COLLINS, G. MUYZER,

and P. WESTBROEK

Abstract. Shell intracrystalline proteinaceous macromolecules isolated from forty four Recent terebratulide
brachiopod species, covering all living superfamilies and two thirds of living families, have been compared
using immunological techniques. Immunological data indicate that the examined species belong to one of the
following four groups, which are also morphologically distinct : ( 1 ) Cancellothyridoidea, (2) ' Terebratelloidea '
(Dallinidae, Terebratelhdae, Laqueidae), (3) Terebratuloidea, and (4) a newly identified category
(Kraussinidae, Megathyrididae, Macandreviidae, Ecnomiosidae). Immunological data clearly indicate that
groups (3) and (4) form a coherent cluster, and that this cluster has a trichotomous relationship with the
remaining two groups. This pattern was not predicted by traditional taxonomies, but reinforced previous
immunological studies. The discovery that Ecnomiosa groups with kraussinids, along with megathyridids and
Macandrevia , allows the presentation of a revised interpretation of terebratulide phylogeny, in which the
extinct zeilleriids and kingenids are considered as a possible link between the Terebratuloidea and the newly
recognized group which includes the Kraussinidae.

The Terebratulida, the largest extant order of brachiopods, is characterized by the presence of a
calcareous loop-shaped brachidium or loop, which functions as a support for the lophophore. The
loop is not only of diagnostic value of the order, but is also of prime taxonomic importance at
almost every rank within the order, because of its highly diversified form and intricate mode of
development. In fact, the loop morphology has been regarded as so important among known
characters, that terebratulide evolution and loop evolution have often been considered as almost
synonymous to each other (Williams and Hurst 1977).

Collins et al. (1988), using immunological methods, demonstrated that the macromolecules
embedded within the microcrystalline calcitic biocrystals of living brachiopod shells contained
significant taxonomic information, and provided a new independent set of characters to reconstruct
terebratulide phylogeny. Subsequent serotaxonomic studies (Collins et al. 1 99 1 zv ; Curry et al.
1991a) indicated that the short-looped superfamily Cancellothyridoidea (Cooper 1973a; ‘-oidea’ is
added to the superfamily stem as the preferred suffix according to the ICZN recommendation. Ride
et al. 1985), the short-looped Terebratuloidea (Cooper 1983), and the long-looped Terebratelloidea
(Muir-Wood et al. 1965) were almost equidistantly related to each other, and that a subset of
the Terebratelloidea was more closely related to the Terebratuloidea than to the rest of the
Terebratelloidea. In effect, this implied that forms with a long loop evolved at least twice
independently.

These first studies of brachiopod serotaxonomy highlighted the problems of existing higher-level
classifications of the Terebratulida, which had been determined only using a few morphological
characters, such as the relative length of the loop and presence or absence of the support for the loop
from the valve floor (Muir-Wood et al. 1965). A considerable revision of traditional classifications
was demanded, as the serotaxonomic data appeared fairly robust, and consistent with the fossil
record (Collins et al. 1991a; Curry et al. 1991a); but a problem remained. From a morphological
standpoint, the relationships suggested by these serotaxonomic studies contained highly improbable
phylogenetic patterns, the prime example being the derivation of the long-looped kraussinids and
Macandrevia from the short-looped Terebratuloidea. This anomaly could not readily be explained

| Palaeontology, Vol. 37, Part 2, 1994, pp. 349-373|

© The Palaeontological Association

350

PALAEONTOLOGY, VOLUME 37

table 1. Samples used in this study. Abbreviations in the ‘Status’ column: L, living when collected and
preserved wet; D, dead shells, preserved dry. * ' Frenulina ’ from off Shimoda, Japan could possibly be a new
species of a new genus. It much resembles Frenulina externally, but the loop is different from that of Frenulina
(D. MacKinnon, pers. comm.).

Species

Locality

Status

Rhynchonellida

Hemithyridae

Notosaria nigricans (Sowerby)

Terebratulida

Terebratulidae

Liothyrella neozelanica (Thomson)

L. uva notorcadensis (Broderip)

Tichosina floridensis Cooper
Gryphus vitreus (Born)

Dyscoliidae

Abyssothyris parva Cooper
Cancellothyrididae
Cancellothyris australis Thomson
Terebratulina retusa (Linnaeus)

T. septentrionalis (Couthouy)

T. unguicula (Carpenter)

T. unguicula rotundata Cooper
T. japonica (Sowerby)

T. peculiaris Hatai
T. pacifica Hatai
T. crossei Davidson
T. reevei Dali

T. abyssicola Adams & Reeve
T. latifrons Dali
T. cailleti Crosse
T. kiiensis Dali & Pilsbry
Chlidonophoridae

Chlidonophora incerta (Davidson)
Dallinidae

Dallina septigera (Loven)

Campages basilanica Dali
Terebratellidae

Terebratella dorsata (Gmelin)

T. sanguined (Leach)

Waltonia inconspicua (Sowerby)
Magellania macquariensis Thomson
Gyrothyris mawsoni antipodensis Foster
Neothyris lenticularis (Deshayes)
Laqueidae

Terebratalia coreaneca (Adams & Reeve)
Coptothyris grayi (Davidson)

Dallinella Occident edis (Dali)

Jolonica nipponica Yabe and Hatai
' Frenulin ’ sp.*

Laqueus rubellus (Sowerby)

L. blanfordi (Dunker)

L. quadrat us Yabe & Hatai
Pictothyris picta (Dillwyn)

Christchurch. New Zealand L

Foveaux Strait, New Zealand L

Ross Island, Antarctica L

off Florida, Gulf of Mexico D

off Corsica, Mediterranean L

off Jacksonville, Florida, USA D

off Melbourne, Australia D

Firth of Lorn, Scotland L

Bay of Fundy, Canada L

Friday Harbor, USA D

off Otsuchi, Japan L

Izu Islands, Japan D

Izu Islands, Japan D

Kii Strait, Japan D

Otsuchi Bay, Japan L

Sibuyan Sea, Philippines D

off Inhambane, Mozambique D

off Key West, Florida, USA D

oft' Pelican Island, Caribbean Sea D

off Valparaiso, Chile D

off Venezuela, Caribbean Sea D

Hebridges Shelf, Scotland L

Izu Islands, Japan D

Strait of Magellan, Argentina D

New Zealand L

Christchurch, New Zealand L

South Pacific D

South Pacific D

Foveaux Strait, New Zealand L

Japan D

Japan L

off San Diego, California, USA D

Izu Islands, Japan D

off Shimoda. Japan D

Sagami Bay, Japan L

Otsuchi Bay, Japan D

Kii Strait, Japan D

Sagami Bay, Japan L

ENDO ET AL.. TEREBR ATULI DE BRACHIOPOD PHYLOGENY

351

TABLE 1. (cont.)

Species

Locality

Status

Macandreviidae

Macandrevia cranium (Muller)

Hebrides Shelf, Scotland

D

M. africana Cooper

off Angola, South Atlantic

D

Ecnomiosidae

Ecnomiosa sp.

Izu Islands, Japan

D

Kraussinidae

Kraussina rubra (Pallas)

Southern Tip, South Africa

D

Megerlia truncata (Gmelin)

off Corsica, Mediterranean

D

Megathyrididae

Megathiris detruncata (Gmelin)

off Corsica, Mediterranean

D

Argyrotheca barretiana (Davidson)

Jamaica, Caribbean Sea

D

in terms of comparative morphology, and prompted misgivings about the immunological approach
(e.g. Brunton and Hiller 1990).

Congruence between molecular and morphological data would provide a strong case for a reliable
interpretation of phylogeny having been established, and hence should be one of the main aims of
any taxonomic study (Curry and Endo 1991). In practice this is particularly desirable for
brachiopod taxonomy, as most species are represented, and often only known, as fossils. In effect,
therefore, suitable morphological explanations are required to justify phylogenetic inferences from
molecular data when there is an apparent discrepancy between molecular and traditional
morphological data. Such explanations are all the more important when there is a risk that the
contrasting interpretations are portrayed as being contradictory and irreconcilable. The reality is
that the majority of relationships suggested by serotaxonomy are entirely consistent with the widely-
accepted classification of brachiopods, and it is the instances where the data are apparently
incompatible that are the focus of attention because they are likely to be critical to the solving of
some of the most puzzling attributes of brachiopod evolutionary history.

The main purpose of this paper, therefore, is to re-examine the contradiction between the
traditional and serotaxonomic views of terebratulide relationships using an enlarged and much
more comprehensive immunological dataset than has previously been available. An attempt is then
made to provide an alternative new phylogenetic interpretation, which may solve the apparent
discrepancy between the serotaxonomic and traditional schemes of phylogenetic interpretations.

MATERIALS AND METHODS

Samples of a total of forty four terebratulide species from world-wide locations were available for
this study (Table 1 ) along with a rhynchonellide species as a control in the immunological assays.
Out of this collection, samples of fifteen different species, representing twelve genera of wide
taxonomic distribution, were available in sufficient abundance to allow the preparation of
polyclonal antisera against shell macromolecules (Table 2). Previous serotaxonomic studies (Collins
et al. 1988 ; Collins et al. 1991a; Curry et al. 1991 a) utilized sub-sets of the antisera used in this study.
The isolation of secondary shell fibres, extraction of intracrystalline macromolecules, immunization,
and immunoassay (ELISA; enzyme-linked immunosorbent assay) procedures follow the methods
described in Collins et al. (1991a), except that, in ELISA, the final detection of the bound antibodies
was performed using a fluorescent substrate [0-2 mM MUP (4-methylumbelliferyl phosphate
dilithium salt; Boehringer Mannheim) in 10 mM diethanolamine buffer containing 2 iiim MgCE
(pH 9-8)], and the fluorescence was read after thirty minutes with a Dynatech Microfluor plate
reader.

352

PALAEONTOLOGY, VOLUME 37

table 2. Antisera used in this study. Abbreviations for the type of antigens: F, extracts from the fibrous
secondary layer; W, extracts from the whole shell; P, semi-purified protein.

Serum
ID No.

Species from which antigen originated

Type of
antigen

Titre

(1/x)

K5038

Notosaria nigricans (Sowerby)

F

3000

802

Liothyrella neozelandica (Thomson)

F

2000

K5010

L. uva notocardensis (Broderip)

F

3000

803

Gryphus vitreus (Born)

W

250

K.4962

Terebratulina retusa (Linnaeus)

W

40000

173

T. septentrionalis (Couthouy)

F

3000

174

T. unguicula (Carpenter)

F

3000

171

T. crossei Davidson

F

400

K5007

Dallina septigera (Loven)

F

10000

K5040

Waltonia inconspicua (Sowerby)

F

20000

427

Neothvris lenticu/aris (Deshayes)

P

100

1 191

Laqueus rubellus (Sowerby)

F

20000

1 192

Pictothyris picta (Dillwyn)

F

5000

801

Kraussina rubra (Pallas)

F

20000

K.5053

Megerlia truncata (Gmelin)

F

8000

Immunological reactivity of antisera was determined for all combinations of antisera and
intracrystalline macromolecules extracted from each species, at an appropriate fixed dilution of each
antiserum. In a series of antiserum dilution assays using the homologous antigen (the antigen
against which the antiserum was prepared), the minimum concentration which gave a 90-100 per
cent reading of the maximum reaction was taken for normal use (Mitre ' in Table 2), while the
concentration which gave 50-70 per cent of the maximum reaction (‘limiting’ concentration) was
used in the inhibition ELISA (see below).

Thirty species, representing all the available genera, were assayed using antisera against twelve
different genera by ELISA to assess the framework of relationships. Cancellothyridid and
chlidonophorid species, a total of fifteen species, were separately assayed by normal ELISA using
four different Terebratulina antisera.

In order to examine in more detail the relationships within each taxonomic group, attempts were
made to increase the specificity of the brachiopod antisera, by mixing, or ‘preabsorbing’, each
antiserum with antigens (shell extracts from species of interest) to remove the antibody reactivity
to that particular antigen (a procedure known as inhibition ELISA; Johnstone and Thorpe 1987).
Preabsorptions were performed either against one and the same antigen for each antiserum simply
to remove common non-informative activities of each antiserum, or against a panel of antigens
extracted from every species within each group to examine detailed inhibition patterns produced by
preabsorptions with different antigens. One part of antigen [in 20 per cent w/v EDTA
(ethylenediamine tetraacetate) solution, pH 8-0] was mixed with nine parts antiserum [in 10 him Tris
buffered saline (TBS, pH 74), 0-02 per cent (v/v) Tween 20, 0-2 per cent (w/v) gelatin solution] with
the appropriate ‘limiting’ concentration of antiserum. Preparations with the homologous antigen
and blank EDTA (20 per cent w/v, pH 8 0) solutions were also included as controls, and were
expected to give 100 per cent and 0 per cent inhibitions respectively. The preparations were
thoroughly mixed, and incubated overnight at 4 °C. The resulting precipitations were removed by
centrifugation at 4000 g for twenty minutes at 4 °C immediately prior to the supernatant being
added to the prepared ELISA plates.

ELISA and inhibition ELISA were performed at least in duplicate. For normal ELISA and
inhibition ELISA using antisera preabsorbed against one and the same antigen, the reactivity of an

ENDO ET AL.\ TEREBRATULIDE BRACHIOPOD PHYLOGENY

353

table 3. Immunological reactivity scores between thirty brachiopod antigens and twelve antisera. Reactions
with homologous antigen are listed as 100, and reactions with negative control (bivalve Codakia sp.) are listed
as 0 Negative readings are listed as 0. Data represents the average of duplicate experiments. Figures with an
asterisk contained 10-20 per cent variations between the duplicate readings, other figures contained less than
10 per cent variations. For antigen identities see Table 1, antiserum identities see Table 2; *1, L. neozelanica ;

*2, L. uva notocardensis'.

*3, T. septentrionalis :

*4, T. dor sat a ; *5, L.

rubellus ; *6,

M. cranium ; *7,

M. africana.

Antisera

5038

803

5010

802

173

5007

5040 427

1191

1 192

801

5053

Antigens

Not

Lio

Lio

Gry

Ter

Dal

Wal Neo

Laq

Pic

Kra

Mer

Rhynchonellida

1 Not os aria

100

0

3

5

*0

15

8 1

0

1

1 1

1

Terebratulida

Terebratuloidea

Terebratulidae

2 Liothyrella *x

0

100

96

134

19

71

16 1

10

41

39

86

3 Liothyrella *2

0

102

100

136

0

64

9 0

10

7

49

89

4 Tichosina

0

92

94

1 19

0

0

8 1

7

0

42

68

5 Gryphus
Dyscoliidae

0

82

80

100

2

0

2 1

6

0

22

40

6 Abyssothyris

0

17

47

0

0

1

9 1

9

12

1 1

1 1

Cancellothyridoidea

Cancellothyrididae

7 Cancellothyris

0

48

9

0

102

104

17

0

7

40

13

8

8 Terebratu/ina *3

0

0

5

0

100

1 16

9

0

7

20

10

5

Chlidonophoridae
9 Chlidonophora

0

4

3

0

90

103

17

0

6

*41

12

7

'Terebratelloidea ’

Dallinidae

10 Dallina

0

0

7

22

1

100

92

1

80

97

16

44

1 1 Campages

0

3

6

0

41

108

96

1

75

98

*19

7

Terebratellidae

12 Terebratella *J

0

8

7

0

0

106

98

70

78

100

22

63

13 Waltonia

0

0

7

1 1

0

105

100

17

88

103

10

7

14 Magellania

0

0

6

0

0

100

95

73

81

101

10

14

1 5 Gyrothyris

0

5

7

0

0

104

100

2

85

102

12

8

16 Neothyris

0

3

6

0

0

103

96

100

84

103

13

5

Laqueidae

1 7 Terebratalia

0

0

7

43

0

113

82

1

78

99

10

9

18 Coptothyris

0

0

6

11

67

1 1 1

87

0

85

102

12

8

19 Dallineila

0

0

3

0

61

108

94

1

77

105

12

7

20 Jolonica

0

8

8

34

0

1 12

93

0

86

101

1 1

7

21 ‘ Frenidina ’

0

5

7

27

41

112

92

1

85

100

12

8

22 Laq ue us *°

0

*4

7

104

0

104

90

I

100

101

12

5

23 Pictothyris

0

8

7

21

0

108

95

0

88

100

12

8

Newly identified category

Macandreviidae
24 Macandrevici *6

0

31

79

105

0

81

58

1 74

86

92

100

25 Macattdrevia*'

0

66

74

*79

0

86

20

1 19

37

*51

74

Ecnomiosidae
26 Ecnomiosa

0

60

89

53

0

86

9

1 10

0

65

85

Kraussinidae
27 Krciussina

0

67

70

94

0

78

27

1 9

27

100

88

0

73

82

124

0

74

9

1 9

4

73

100

28 Megerlia
Megathyrididae
29 Megathiris

0

33

49

30

0

*84

*10

1 29

57

50

79

30 Argyrotheca

0

47

6

0

0

1

6

1 1 1

0

13

7

354

PALAEONTOLOGY, VOLUME 37

text-fig. 1. (7-mode bivariate comparisons of immunological reactivity data. Similarity coefficients of cosine 6
measure (cf. Lesperance 1990) between each pair of antigens (rows in Table 3) were calculated, and then-
values indicated as the interval to which each value belonged. High values indicate high similarities. Figures
along x- and r-axes correspond with the species (antigen) identity numbers in Table 1. Four major groups
(Tu, C, Te, K) have been recognized within the Terebratulida, demonstrating high similarities within each

group and between the groups Tu and K.

antiserum to an unknown antigen was given as a percentage of the fluorescence reading of the
sample to that of the positive control (i.e. reaction with homologous antigen). In both cases the
negative control reading (i.e. reaction with shell extracts from the bivalve Codakia sp.) was
subtracted before the calculation. In inhibition ELISA using antisera preabsorbed against a panel
of different antigens, the reactivity of an antigen to each preabsorbed antiserum was also expressed
as a percentage, where the reaction with the antiserum preabsorbed by the homologous antigen, and
the reaction with the non-preabsorbed antiserum ('preabsorbed' with EDTA) were taken as 0 and
100 per cent, respectively.

The resulting immunological reactivity data were transformed to distance matrices, expressing
the degree of similarity both among antigens ((7-mode) and antisera (/Tmode), using the similarity
coefficients of cosine 9 measure and Euclidean distance (cf. Lesperance 1990), to produce
dendrograms by the single linkage clustering methods (cf. Sneath and Sokal 1973). Data analyses
were performed with the aid of an Apple Macintosh computer using the Odesta Corporation
package 'Datadesk Professional version 2.0’.

RESULTS

Framework relationships of the Terebratulida

The calculated reactivity scores for thirty antigens of the examined genera against twelve
brachiopod antisera are summarized in Table 3. The rhynchonellide species, included as a control
to check the specificity of the antisera, is clearly discriminated from the terebratulides. The
antiserum (laboratory number K5038) prepared against the rhynchonellide Notosaria reacted only
with Notosaria antigen and did not react with any others. The Notosaria antigen (i.e. the
macromolecules extracted from the shell of Notosaria ) showed no significant reactions with any of
the antisera, except with K5038 itself.

ENDO ET AL.: TEREBRATULIDE BRACHIOPOD PHYLOGENY

355

table 4. Inhibition ELISA on Tu and K groups (Terebratulidae, Macandreviidae, Ecnomiosidae, Kraussinidae
and Megathyrididae). Each of four different antisera (803, 802, 801, K5053) was preabsorbed with each of ten
different antigens (columns) from species of Tu and K groups, and assayed with the ten antigens (rows).
Reaction with the non-preabsorbed antiserum are listed as 100. while those with the antiserum preabsorbed
by the homologous antigen are listed as 0. Data represent the average of duplicate experiments.

Antisera (preabsorbed against)

Antigens

-Ln

-Lu

-Gv

-Tf

-Md

-Me

-Ma

-Ec

-Kr

-Mt

Anti-Liothyrella (803)

Liothyrella neozelanica (Ln)

0

0

21

19

100

91

89

100

85

88

Liothyrella uva (Lu)

0

0

0

94

100

100

87

100

87

45

Gryphus vitreus (Gv)

0

0

0

3

92

97

22

74

79

88

Trichosina floridensis (Tf)

0

0

16

19

91

90

80

91

70

86

Megathiris detnmcata (Md)

0

0

2

0

3

0

3

0

1

0

Macandrevia cranium (Me)

0

0

7

0

10

0

10

0

8

0

Macandrevia africana (Ma)

0

0

1

0

71

56

7

82

11

55

Ecnomiosa sp. (Ec)

0

0

1

0

6

5

6

21

2

0

Kraussina rubra (Kr)

0

0

0

0

61

56

40

83

i

25

Megerlia truncata (Mt)

0

0

0

0

36

56

43

69

9

6

Anti-Gryphus (802)

Liothyrella neozelanica

0

0

0

0

100

98

49

100

62

58

Liothyrella uva

0

0

0

0

96

81

55

95

50

55

Gryphus vitreus

0

0

0

8

89

90

0

0

66

76

Tichosina floridensis

0

0

0

0

89

97

70

100

52

78

Megathiris detruncata

0

0

0

0

63

53

27

79

3

5

Macandrevia cranium

0

0

0

0

59

0

37

64

26

0

Macandre via africana

0

0

0

0

81

58

42

77

19

6

Ecnomiosa sp.

0

0

0

0

70

52

39

73

7

2

Kraussina rubra

0

0

0

0

100

69

68

100

29

36

Megerlia truncata

0

0

0

0

83

23

47

93

33

0

Amti-Kraussina (801)

Liothyrella neozelanica

2

0

2

0

0

0

0

0

0

0

Liothyrella uva

14

0

5

0

9

0

8

3

0

0

Gryphus vitreus

3

0

2

0

3

0

0

0

0

0

Tichosina floridensis

2

0

3

1

5

1

4

3

0

0

Megathiris detruncata

i

1

0

0

4

2

1

3

0

3

Macandrevia cranium

75

41

60

51

0

0

0

0

0

0

Macandre via africana

18

8

15

28

0

0

0

2

0

0

Ecnomiosa sp.

20

20

31

23

2

1

2

i

0

1

Kraussina rubra

83

93

94

91

56

22

63

79

0

79

Megerlia truncata

41

26

49

56

6

0

4

6

0

0

Anti-Megerlia (K5053)

Liothyrella neozelanica

9

21

48

46

36

51

54

68

39

0

Liothyrella uva

31

0

38

15

31

0

32

22

1 1

0

Gryphus vitreus

0

0

5

6

6

0

0

0

0

0

Tichosina floridensis

10

0

15

23

24

1 1

35

45

0

0

Megathiris detruncata

56

58

63

72

2

13

37

21

3

0

Macandrevia cranium

69

46

70

76

13

0

15

18

0

0

M acandrevia africana

69

54

86

82

43

0

24

41

1 1

0

Ecnomiosa sp.

79

67

84

79

20

9

36

31

0

0

Kraussina rubra

80

69

83

85

37

26

57

46

7

0

Megerlia truncata

85

80

90

89

61

67

72

79

58

0

356

PALAEONTOLOGY, VOLUME 37

discrimination of the two groups. Euclidean distance was used as the similarity coefficient (cosine 0 measure
could not be calculated because of the 0 values in the data), a, data from assays using anti -Liothyrella (803)
serum; b, using anti -Gryphus (802); c, using anti- Kraussina (801); d, using anti -Megerlia (K5053). For species

identity, see Table 4.

Antisera prepared against Liothyrella , Gryphus , Kraussina, and Megerlia (803, K5010, 802, 801,
K5053) all showed very similar patterns of reactivity, with strong reactions with the terebratulids,
kraussinids, megathyridids, Macandrevia , and Ecnomiosa , and little reaction with other species. The
dyscoliid Abyssothyris and the megathyridid Argyrotheca were among the least reactive terebratulide
antigens to most of the antisera, but reacted moderately with one of the two antisera prepared
against Liothyrella (803, K5010).

Antiserum prepared against Terebratulina septentrionalis (173) (the antisera are hereafter
described as ‘anti -Terebratulina septentrionalis', etc.) reacted strongly with cancellothyridids and
Chlidonophora , and showed weaker reactions with only a few dallind and laqueid species.
Cancellothyridid and chlidonophorid antigens reacted strongly with anti -Terebratulina (173) and
anti -Dallina (K5007), and moderately with anti -Pictothyris (1 192).

Antisera prepared against Waltonia, Laqueus, and Pictothyris (K5040, 1191, 1 192) again had very
similar reaction profiles, showing strongest reactions with terebratellids, dallinids and laqueids,
moderately strong reactions with one of the Macandrevia species (M. cranium), and sporadically
weaker reactions with other terebratulides. Anti -Dallina (K5007) was the least specific antiserum,
but it barely reacted with the rhynchonellide, and only weakly with terebratulids, Abyssothyris ,
kraussinids, megathyridids, Macandrevia , or Ecnomiosa. Anti- Neothvris (427) serum, on the
contrary, reacted very specifically only with the terebratellid genera, Neothyris, Magellania,
Terebratella , and Waltonia. These patterns of reactions, as described above, were visualized by a
pair-wise comparison of the antigens (Text-fig. I).

These data reinforce the three-fold division of living terebratulides proposed by previous
immunological investigations (Collins et al. 1991c/; Curry et at. 1991c/), and assign the examined
taxa into the following three divisions, involving four major taxonomic units.

ENDO ET AL.: TEREBRATULIDE BRACHIOPOD PHYLOGENY

357

table 5. Inhibition ELISA on Te group (Terebratellidae, Dallinidae, and Laqueidae). Each of four different
antisera (K5007, K.5040, 427, 1191) was preabsorbed with each of ten different antigens (columns) from species
of Te group, and assayed with the ten antigens (rows). Reactions with the non-preabsorbed antiserum are listed
as 100, while those with the antiserum preabsorbed by the homologous antigen are listed as 0. Data represent
the average of duplicate experiments.

Antisera (preabsorbed against)

Antigens

-Nl

-Td

-Ts

-Wi

-Mm

-Gm

-Ds

-Cb

-Lr

-Do

Anti-Dallina (K5007)

Neothyris lenticularis (Nl)

0

ft

0

16

8

10

0

66

60

49

Terebratella dorsata (Td)

0

ft

5

27

10

21

0

70

64

59

Terebratella sanguined (Ts)

0

ft

0

0

8

3

0

69

0

53

Waltonia inconspicua (Wi)

0

0

0

6

6

5

0

69

59

53

Magellania macquariensis (Mm)

0

0

ft

14

6

15

0

64

62

49

Gyrothyris mawsoni (Gm)

0

ft

0

6

7

10

0

63

58

50

Dallina septigera (Ds)

0

0

7

25

10

26

ft

69

69

56

Campages basilanica (Cb)

0

ft

0

10

4

9

0

60

53

37

Laqueus rubellus (Lr)

0

ft

0

0

4

7

0

62

33

27

Dallinella occidentalis (Do)

0

ft

0

ft

8

4

0

64

43

34

Anti-Waltonia (K5040)

Neothyris lenticularis

0

20

21

ft

9

0

0

41

3

ft

Terebratella dorsata

0

8

2

0

13

10

7

41

14

27

Terebratella sanguined

0

29

48

ft

1 1

ft

ft

57

0

29

Waltonia inconspicua

0

24

8

ft

18

8

13

55

20

26

Magellania macquariensis

0

14

6

0

9

0

5

47

6

12

Gyrothyris mawsoni

4

30

12

0

19

0

0

36

12

24

Dallina septigera

0

9

28

0

5

ft

0

56

14

18

Campages basilanica

2

1

0

0

0

0

0

0

ft

0

Laqueus rubellus

0

21

9

ft

12

ft

0

39

0

0

Dallinella occidentalis

0

21

1

0

17

ft

0

28

ft

ft

Anti-Neothvris (427)

Neothyris lenticularis

0

33

83

75

32

58

65

82

100

90

Terebratella dorsata

0

ft

68

75

0

79

45

76

93

98

Terebratella sanguined

0

0

0

93

ft

22

1

39

65

50

Waltonia inconspicua

0

7

44

1 1

13

54

50

64

57

32

Magellania macquariensis

0

7

69

77

ft

73

48

76

87

72

Gyrothyris mawsoni

0

1

0

ft

4

1

2

15

28

17

Dallina septigera

0

ft

21

53

0

40

4

57

62

69

Campages basilanica

0

4

0

ft

0

0

ft

0

26

7

Laqueus rubellus

0

0

0

0

0

6

ft

1

ft

8

Dali inella occidental is

ft

0

15

0

0

29

23

9

16

5

Anti- Laqueus (1191)

Neothyris lenticularis

0

9

ft

ft

7

10

8

37

ft

17

Terebratella dorsata

0

2

0

0

4

5

6

35

ft

15

Terebratella sanguined

0

22

ft

0

4

2

2

34

ft

21

Waltonia inconspicua

0

10

ft

0

6

6

4

41

ft

23

Magellania macquariensis

ft

0

ft

ft

ft

2

3

31

0

1 1

Gyrothyris mawsoni

ft

0

1

ft

1

2

2

27

ft

1 1

Dallina septigera

0

1

ft

0

5

4

3

32

0

16

Campages basilanica

0

0

ft

0

ft

ft

1

4

ft

ft

Laqueus rubellus

59

77

78

80

79

75

81

81

ft

69

Dallinella occidentalis

34

38

38

41

40

44

49

53

0

23

358

PALAEONTOLOGY, VOLUME 37

table 6. Inhibition ELISA on Te group using antisera preabsorbed with Macandrevia antigen. Each of five
different antisera (K5007, K5040, 427, 1191, 1 192) was preabsorbed with antigens from Macandrevia cranium
(columns), and assayed with the 15 antigens from Te group species (rows). Reactions with the homologous
antigen are listed as 100, while reactions with the negative control (bivalve Codakia sp.) are listed as 0. Data
represent the average of duplicate experiments.

Antigens

Antisera

K5007

(Dal)

K5040

(Wal)

427

(Neo)

1 191
(Laq)

1192

(Pic)

Dallina

(Ds)

100

43

35

75

98

Campages

(Cb)

115

88

18

76

72

Gyrothyris

(Gm)

144

101

17

79

75

Waltonia

(Wi)

138

100

26

77

101

Terebratella dorsata

(Td)

109

77

87

76

85

Neothyris

(Nl)

96

96

100

73

97

Jolonica

(Jn)

0

96

0

72

97

' Frenulina '

(Fr)

35

1 16

0

76

111

Terebratalia

(Tc)

51

1 10

0

103

101

Coptothyris

(Cg)

57

108

0

92

110

Dallinella

(Do)

68

1 10

0

85

104

Laqueus rubellus

(Lr)

17

96

0

100

84

L. blanfordi

(Lb)

81

90

0

94

94

L. quadratus

(Lq)

53

77

0

86

85

Pictothyris

(Pp)

10

118

0

51

100

Mm

Td

Nl

Dl

Lr

te

la

Cb

Gm

Ts

Ds

Wi

te

+

da

text-fig. 3. R-mode cluster analyses on inhibition data on Te groups (Table 5), demonstrating a general
separation of terebratellids and laqueids. Euclidean distance was used as the similarity coefficient, a, data from
assays using anti -Dallina (K5007) serum; b, using anti- Waltonia (K5040); c, using anti -Neothyris (427);
d, using anti -Laqueus (1 191). For species identity, see Table 5.

ENDO ET A L. \ TEREBRATULIDE BRACHIOPOD PHYLOGENY

359

text-fig. 4. (7-mode cluster analysis on
inhibition data on Te group (Table 6),
demonstrating a clear discrimination
between terebratellids and laqucids.
Euclidean distance was used as the simi-
larity coefficient. For species identity, see
Table 6.

text-fig. 5. (7-mode cluster analysis on
immunological reactivity data on C
group (Table 7). Antisera were not
preabsorbed. Euclidean distance was
used as the similarity coefficient. For
species identity, see Table 7.

(1) Short-looped Terebratuloidea Cooper, 1983, denoted as Tu group here, plus a group of long-
looped families Kraussinidae, Megathyrididae, Macandreviidae Cooper, 19736, and Ecnomiosidae
Cooper, 1977 (collectively referred to here as the K group).

(2) Short-looped Cancellothyridoidea Cooper, 1 973c/ the C group.

(3) A group of long-looped families Dallinidae, Laqueidae (Richardson 1975; except for
Macandrevia), and Terebratellidae - the Te group (Text-fig. 1 ).

Strong and consistent correlations between Tu and K groups were demonstrated, while no affinity
between the short-looped Tu and C groups, and only weak affinities between the long-looped Te and
K groups were observed (Text-fig. 1 ), confirming the critical disagreements with traditional
classifications suggested by the original immunological study (Collins et al. 1988).

Relationships within major taxonomic groups

These results, based entirely on assays carried out using crude antisera, clearly separated the three
major groups, but the data were generally insufficient to resolve relationships within each group as
many of the reactions were oversaturated (Table 3). Separation between the morphologically
distinctive Tu and K groups was also unclear (Text-fig. 1). However the more specific inhibition
ELISA on the species belonging to Tu and K groups clearly discriminated between these two groups
(Table 4; Text-fig. 2). Coherence of the genus Liothyrella and family Kraussinidae was also
suggested (Text-fig. 2a), but generally the data were still too noisy to allow further elucidation of
relationships. Assays on species of the Te group using the preabsorbed antisera separated the
‘laqueids’ of the northern hemisphere and the terebratellids of the southern hemisphere, but failed
to detect the ‘dallinids’ ( Dallina and Campages) as a coherent group (Tables 5-6; Text figs 3-4).
More work needs to be carried out to determine the relationships of dallinids to the other two

360

PALAEONTOLOGY. VOLUME 37

between r and u subgroups, and separation of these subgroups from 77 crossei and 77 pacifica. Euclidean
distance was used as the similarity coefficient, a, data from assays using anti-Terebratulina retusa (K4962)
serum; B, using anti- 77 septentrionalis (173); C, using anti- 77 unguicula ( 1 74) ; d, using anti- 77 crossei (171). For

species identity, see Table 8.

table 7. Immunological reactions among C group species (Cancellothyrididae and Chlidonophoridae).
Reactions with the homologous antigen are listed as 100, while reactions with the negative control (bivalve
Codakia sp.) are listed as 0. For details of the antisera and antigens, see Tables 1 and 2. Antisera were not
preabsorbed with antigens.

Antigens

Antisera

K.4962

(ret)

173

(sep)

174

(ung)

171

(cro)

Terebratulina retusa

(ret)

100

90

87

83

T. septentrionalis

(sep)

102

100

97

105

T. unguicula

(ung)

93

87

100

77

T. unguicula rot undata

(rot)

93

90

102

61

T. japonica

(jap)

106

89

93

95

T. peculiar is

(pec)

107

95

98

99

T. pacifica

(pac)

56

81

78

61

T. crossei

(cro)

59

75

66

100

T. reevei

(ree)

56

29

72

51

T. abyssicola

(aby)

75

44

76

11

T. latifrons

(lat)

73

36

92

67

T. cailleti

(cai)

92

63

93

29

T. kiiensis

(kii)

46

66

77

96

Chlidonophora incerta

(Chi)

99

86

102

73

C an cello thy r is australis

(Can)

1 13

94

105

60

families. The ‘laqueids’ group comprises Laqueus , Pictothyris , Jolonica, Terebratalia , Coptothvris ,
and Dallinella (Text-lig. 4), and hence the immunological data can he considered as supporting the
assignment of these genera to the Laqueidae (Richardson 1975).

ENDO ET A L TEREBRATULIDE BRACHIOPOD PHYLOGENY

361

table 8. Inhibition ELISA on Terebratulina species. Each of four different antisera (K4962. 1 73. 174. 171) was
preabsorbed with each of eight different antigens (columns) from Terebratulina species (C group), and assayed
with the eight antigens (rows). Reactions with the non-preabsorbed antiserum are listed as 100. while those
with the antiserum preabsorbed by the homologous antigen are listed as 0. Data represent the average of
duplicate experiments.

Antigens

Antisera (preabsorbed against)

-re

-se

-un

-ro

-ja

-pe

-pa

-cr

Anti-T. retusa (K4%2)

Terebratulina retusa (re)

0

7

22

25

16

0

33

47

T. septentrionalis (se)

0

0

14

22

1

0

15

18

T. unguicula (un)

0

4

0

0

5

4

40

50

T. unguicula rotundata (ro)

0

5

0

0

1

2

39

24

T. japonica ( j a )

0

7

10

5

0

6

36

44

T. peculiaris (pe)

0

13

19

18

3

ii

42

53

T. pacifica (pa)

0

0

1 1

12

6

3

(1

1 1

T. crossei (cr)

0

1)

0

3

0

2

0

0

Anti-T. septentrionalis (173)

Terebratulina retusa

5

0

32

29

14

40

57

60

T. septentrionalis

0

11

27

37

19

26

52

69

T. unguicula

8

1)

20

34

22

35

64

83

T. unguicula rotundata

8

0

24

31

12

31

56

78

T. japonica

10

0

40

48

10

33

59

73

T. peculiaris

0

0

38

38

12

13

65

84

T. pacifica

10

0

33

35

3

8

19

50

T. crossei

1 1

I)

21

35

0

(1

16

38

Anti-T. unguicula (174)

Terebratulina retusa

20

31

0

1

5

9

76

95

T. septentrionalis

37

27

0

1

2

19

75

79

T. unguicula

45

45

0

0

47

40

70

86

T. unguicula rotundata

45

43

0

3

42

48

80

87

T. japonica

50

41

0

2

13

15

78

92

T. peculiaris

32

31

0

0

13

8

77

81

T. pacifica

17

7

0

1

(1

1

27

59

T. crossei

21

5

0

2

13

15

43

14

Anti-T. crossei (171 )

Terebratulina retusa

0

0

87

97

0

0

0

0

T. septentrionalis

((

0

50

80

4

0

19

0

T. unguicula

23

12

8

41

18

43

76

0

T. unguicula rotundata

9

0

0

21

25

15

46

0

T. japonica

1)

(1

62

70

0

0

0

0

T. peculiaris

0

0

85

76

19

0

4

0

T. pacifica

0

0

70

92

0

0

0

0

T. crossei

32

27

72

88

47

19

31

1)

Antisera prepared against four Terebratulina species, even without preabsorption treatment,
detected a considerable amount of molecular variation among C group species (Table 7).
The variability between species of Terebratulina was even greater than that detected between the
families of Te group. C group species were subdivided into two major groups; one comprised
T. retusa , T. septentrionalis, T. japonica , T. peculiaris , T. unguicula , T. unguicula rotundata ,
Cancellothyris australis , and Chlidonophora incerta , the other group comprised T. pacifica , T. reeve i.

362

PALAEONTOLOGY, VOLUME 37

Group Family

Loop morphology

Terebratulina retusa
T. septentrionalis
T. japonica
T. peculiaris
T. unguicuia
T. unguicuia rotundata
Cancellothyris australis
Chlidonophora incerta
T. pacifica
T. cailleti
T. reevei
T. latifrons
T. kiiensis
T. crossei
T. abyssicola

Campages basilanica
Dallina septigera
Neothyris lenticularis
Terebratella dorsata
Magellania macquariensis
Waltonia inconspicua
Gyrothyris mawsoni
Terebratella sanguinea
Laqueus rubellus
L. blanfordl
L. quadratus
Dallinella occidentalis
Coptothyris grayl
Terebratalla coreanica
"Frenulma" sp.

Jolonica nipponlca
Plctothyris pi eta

Liothyrella neozelanica

L. uva notorcadensis
Gryphus vitreus
Tichosina floridensis
Abyssothyris parva

Argyrotheca barretiana
Megathiris detruncata
Macandrevia cranium

M. africana
Ecnomiosa sp
Kraussina rubra
Megerlia truncata

Notosaria nigricans

ca

Terebratulina

text-fig. 7. Summary of immunological view of terebratulide relationships, with stylized loop morphology
(after Davidson 1886, except for the original Ecnomiosa) of selected genera. Polychotomous branchings
indicate unresolved relationships. Horizontal and vertical axes are not proportional to any similarity values.
Abbreviations for families: ca, Cancellothyrididae; ch. Chlidonophoridae; da, Dallinidae; te, Terebratellidae;
la, Laqueidae; tu, Terebratulidae; dy, Dyscoliidae; me, Megathyrididae; ma, Macandreviidae;

ec. Ecnomiosidae; kr, Kraussinidae.

ENDO ET AL. TEREBRATULI DE BRACHIOPOD PHYLOGENY

363

T. crossei , T. kiiensis, T. latifrons , T. cailleti , and T. abyssicola (Text-fig. 5). The data further suggest
the separations of the following subgroups: T. crossei and T. kiiensis (subgroup c); and T. latifrons
and T. reevei (subgroup /) (Text-fig. 5). Assays with preabsorbed antisera confirmed the separation
between subgroup r ( 7 . retusa, T. septentrionalis, T. japonica, and 7 . peculiaris) and subgroup u
(T. unguicula , T. unguicula rotundata ) (Table 8; Text-fig. 6). The apparent affinity of both
Cancellothyris and Chlidonophora to T. unguicula and T. unguicula rotundata (Table 7; Text-fig. 5)
suggests that the diversification of the ancestors of the living Terebratulina occurred before the
divergence of Chlidonophora and Cancellothyris from Terebratulina stocks.

The relationships among the other cluster of Terebratulina species were poorly characterized
because anti-7", crossei was the only antiserum directed to a species belonging to this cluster. This
antiserum indicated that T. crossei was very distantly related to T. abyssicola and T. cailleti (Table
7; Text-fig. 5). The assays with preabsorbed antisera demonstrated a large molecular variation
between T. crossei and T. pacifica (Table 8). These facts suggested an involvement of a number of
deeply branched lineages in this second major group of Terebratulina.

A schematic summary of the immunological results concerning the relationships among examined
terebratulide species is given in Text-figure 7.

DISCUSSION

Nature of the immunological data

Immunological methods utilize the highly specific reaction between antigenic molecules, commonly
proteins or polysaccharides, and antibodies. The production of the latter is elicited by the
introduction of antigens into the body fluids of a susceptible animal (typically, a rabbit). The degree
of reaction may vary depending on the structural similarity between the assayed molecules and
those that originally elicited the antibodies. When applied to phylogenetic studies, immunological
data therefore provide a measure of overall similarities between examined biomolecules; this then
can be extrapolated to the degree of divergence between organisms that carry the molecules. The
biomolecules thus compared in this study were the brachiopod intracrystalline macromolecules
occluded in the secondary shell layer.

Compositional analysis has demonstrated that the brachiopod secondary shell fibrous calcites
contain proteins, lipids (Curry et at. 19916), and neutral carbohydrates (Collins et ah 19916; Clegg
and Moers, pers. comm.). The brachiopod intracrystalline proteins have been partially characterized
by amino acid analysis, gel electrophoresis, liquid chromatography, and N-terminal amino acid
sequencing (Curry et al. 1991 6, 1991c; Collins et ah 1 99 1 b ; Cusack et al. 1992). Curry et al. ( 1991c)
reported that at least three different proteins of discrete sizes (47 kDa, 16 kDa, and 6-5 kDa) are
present in the terebratellid Neothyris lenticularis (Deshayes) and at least one protein (30 kDa) in the
cancellothyridid Terebratulina retusa (Linnaeus). The amino acid sequences of these proteins had
no significant similarity with known proteins (Curry et al. 1991c). The functions of brachiopod
intracrystalline proteins are poorly known, although one protein is believed to be responsible for
shell colour (Curry et ah 19916; Cusack et ah 1992). Our immunological data did not group taxa
of a particular shell colour; therefore, it appears very unlikely that the colour-related proteins
evolved convergently to produce the present patterns of immunological reactivities among
terebratulide species.

The antibodies utilized in this study and previous studies (Collins et al. 1988 ; Collins et cd. 1991 a:
Curry et al. 1991c/) were prepared against crude extracts from the fibrous secondary layer of the shell
or the whole shell powder, except for the anii-Neothyris serum which was prepared against a semi-
purified protein (Collins et al. 1991cv). Immunological assays using these sera on the liquid
chromatography fractions of purified shell extracts indicated that these antisera were directed not
only against proteins but also against carbohydrates (Collins et al. 19916; Endo, unpublished data).

The nature and extent of glycosylation in the brachiopod intracrystalline proteins remains
unclear. Heavy glycosylation was suggested by Collins et ah (19916). The fact that the first twenty
amino acids of the ' 1 0-5 kDa protein' and the first ten of the ‘47 kDa protein’ could be sequenced

text-fig. 8. Interpretations of terebratulide phylogeny. Only relevant taxa are indicated. Upper panel:
traditional interpretations, a, Muir-Wood et at. (1965) with modifications by Cooper (1973a, 1979):
b. Smirnova (1984). Lower panel, interpretations based on immunological data, c, Collins el al. (1991a): D, the
revised interpretation.

PALAEONTOLOGY. VOLUME 37

364

PALAEONTOLOGY, VOLUME 37

A B

321"

P

Te

Tu K

I I

\

v .2.

Palaeozoic ancestors

text-fig. 8. Interpretations of terebratulide phytogeny. Only relevant taxa are indicated. Upper panel:
traditional interpretations, a, Muir-Wood et al. (1965) with modifications by Cooper (1973a, 1979);
b. Smirnova (1984). Lower panel, interpretations based on immunological data, c, Collins et al. (1991a); D, the

revised interpretation.

ENDO ET A L. TEREBRATULIDE BRACHIOPOD PHYLOGENY

365

with the Edman degradation (Curry et al. 1991c) indicates that at least these residues have no
covalently hound sugars. Curry et al. (1991/?) reported no galactosamine or glucosamine from the
brachiopod shell fibre extracts.

In general, the immunological data were most consistent and unequivocal at family-superfamily
levels and above. The long-lived (Jurassic-Recent) genus Terebratulina , for which species level
inferences were possible, was a remarkable exception. Indeterminate reaction patterns observed at
levels lower than family rank may be due partly to the fact that antibodies were directed also against
carbohydrate epitopes, which are usually considered as much less informative than protein epitopes
(Cohen 1992). Another possible cause for spurious immunological reaction patterns is the variation
in the antigen concentration per shell weight, since the amount of antigen was adjusted by the
amount of dried shell fibres in this study. This factor may explain minor variations, for example,
the slightly odd reaction patterns of Gryphus vitreus and Macandrevia cranium antigens, which gave
systematically weaker and stronger reactions, respectively, compared with other antigens (Table 3).

In this study, we use common similarity coefficients and simple clustering algorithms to
graphically present structures of the immunological data. Different tree-building methods may
produce minor changes in the tree topologies, but the framework of relationships, the separation
of Tu , C, Te, and K groups and their interrelationships, will not change, as the raw data set (Table
3) almost self-evidently demonstrates.

Traditional interpretations of terebratulide phytogeny

In Muir-Wood et al. (1965), post-Palaeozoic terebratulides were assigned in one of the two
suborders, the short-looped Terebratulidina and the long-looped Terebratellidina. The former
consisted of the single superfamily Terebratuloidea, and the latter embraced two superfamilies,
namely the Terebratelloidea and the extinct Zeillerioidea. Dagys (1968, 1972) proposed a different
subdivision of the long-looped terebratulides, erecting the new superfamily Dallinoidea, which
embraced the Zeilleriidae, Dallinidae and Laqueidae on the basis of the presence or absence of
dental plates and a cardinal process, in addition to loop characteristics. Dagys (1968, 1972)
concluded that the other long-looped superfamily Terebratelloidea, consisting of the Terebratellidae,
Megathyrididae, Platidiidae and Kraussinidae, was derived not from the Dallinoidea but from the
newly erected short-looped superfamily Loboidothyroidea, casting doubt on the assumption implicit
in Muir-Wood et al. (1965) interpretation, or the separation of the long- and short-looped forms
at subordinal rank.

Among the short-looped Terebratuloidea, Cooper (1973c/) recognized in the Cancellothyrididae
a fundamental difference in the way the pedicle muscles attach to the dorsal shell. In particular he
noticed that the muscles attach to the valve floor rather than to the hinge plate, which is unlike the
situation in other terebratuloids. On the basis of this muscle attachment and the characteristic
features of the cardinalia. Cooper ( 1 9 7 3 <r/ ) erected the superfamily Cancellothyridoidea. The long
looped superfamily was also divided into two superfamilies, the Terebratelloidea and Dallinoidea
(e.g. Cooper 1979), presumably after Dagys (1972), but the separation of the long- and short-looped
suborders (Cooper 1 98 1 c/, 19816, 1982) and the separation of the superfamily Zeillerioidea from
other long looped superfamilies (Cooper 1989), were retained (Text-fig. 8a). As to the relationships
among long-looped terebratulides, Richardson (1975) regarded the following three families,
Terebratellidae, Dallinidae, and Laqueidae as different from the rest of the then known long-looped
forms (the Kraussinidae, Platidiidae, Megathyrididae, and Thaumatosiidae) in many respects,
considering that the former three families constitute the typical stock of the superfamily
Terebratelloidea.

In the latest account of the traditional classification of brachiopods, Smirnova (1984) followed
Dagys (1968, 1972) as to both the separation of the Terebratelloidea and Dallinoidea, and the
connection of the Terebratelloidea with the short-looped Loboidothyroidea, but with major
amendments concerning the origins of these superfamilies (Text-fig. 8b). The familial separations of
the Dallinidae, Laqueidae, Macandreviidae, Kingenidae and the new family Diestothyrididae were

366

PALAEONTOLOGY, VOLUME 37

recognized, and they were included in the superfamily Dallinoidea, distinguishing this superfamily
from the other long-looped superfamily Terebratelloidea which embraced the Terebratellidae,
Megathyrididae, Platidiidae, and Kraussinidae (Smirnova 1984).

Relationships demonstrated by the immunological data

Immunological data clearly indicate that the long loop of the living terebratulides evolved at least
twice (Text-fig. 7). This pattern of relationships contradicts not only the interpretations of Muir-
Wood et al. and Cooper (Text-fig. 8a) but also the scheme of Dagys and Smirnova (Text-fig. 8b).
This result was surprising, but may not be as radical a suggestion as it may at first appear.
Significantly, perhaps, every one of the new long-looped terebratulide taxa recognized in this study
- kraussinids, megathyridids, Macandrevia , and Ecnomiosa - has characteristic ‘aberrant’ mor-
phological features which distinguish it from ‘typical’ long-looped terebratulides, such as the
dallinids, terebratellids and laqueids. The kraussinids and megathyridids have a simpler structure
of the lophophore and loop, which is believed to be neotenously derived from the typical long-
looped stocks (Williams and Hurst 1977). The adult loop of Macandrevia is not particularly
aberrant, but the cardinalia have very characteristic features which are reflected in the various
assignments of this genus to the Dallinidae (Muir-Wood et al. 1965), Macandreviidae (Cooper
19736), and Laqueidae (Richardson 1975). The adult loop of Ecnomiosa , on the other hand, is
superficially also of standard long-looped type. However the mode of loop development is
distinctive in this genus, and aspects of its cardinalia are unique. Cooper (1977, p. 129) considered
that ‘the loop of this brachiopod is so unusual as to set the genus apart from all others known’, and
erected the new family Ecnomiosidae for this genus, a family which was later assigned to the
Dallinoidea (Cooper 1 98 1 tv).

Another major finding of the earlier immunological investigations that was confirmed in this
study was that the Cancellothyridoidea (Cooper 1973c/; C group. Text-fig. 9) was almost
equidistantly related to both a subset of the Terebratelloidea (Muir-Wood et al. 1965; Te group)
and the Terebratuloidea (Cooper 1983; Tu group) to which another subset of the Terebratelloidea
(A' group) was linked. From the earliest fossil record of these three major superfamilies and the
families to which the second long-looped species belong, Collins et al. (1991 a) and Curry et al.
(1991u) postulated that the last common ancestor of all living terebratulides diverged in the Triassic,
and Collins et al. (1991/7) further suggested that the second long-looped lineage diverged earliest in
the Cretaceous from the terebratuloid stocks (Text-fig. 8c). These interpretations are not
inconsistent with available information from the fossil record, at least in respect of the time of the
first occurrence for each relevant taxon. But, considering the fact that the Terebratuloidea had
already been morphologically established in the Cretaceous (Cooper 1983), the number of
simultaneous mutations required to derive the megathyridids in the Cretaceous, or kraussinids in
the Tertiary from terebratuloids, are very large, as Brunton and Hiller (1990) have pointed out; and
the relationships among the second long-looped species were virtually unknown.

The questions are, therefore, how the species of the second long-looped terebratulides recognized
in this study (A' group) can be related to each other - the morphological affinities among the other
long-looped group (TV group), dallinids, terebratellids, and laqueids, have been suggested by
Richardson (1975) -and how the second long-looped lineage could be derived from the short-
looped Terebratuloidea? The key seems to lie with the ‘aberrant’ loop and cardinalia of Ecnomiosa
and Macandrevia.

Morphology of Ecnomiosa and Macandrevia

The specimen of Ecnomiosa used in the immunological assay was one of the two adult specimens
collected from a locality in Japanese waters. It was a dead, though fresh, complete articulated shell
(the other being alive when collected) which allowed observations of both the external features and
the internal features, including the loop, and unequivocal identification of the genus. The size, the
extent of sulcation, and the geographical distribution suggest that these individuals constitute a new

ENDO ET AL T E R E B R ATU L I D E BRACHIOPOD PHYLOGENY

367

text-fig. 9. Dorsal valve interior morphology of Ecnomiosa , Macandrevia , and Zeilleria. A, Ecnomiosa
inexpectata , a juvenile specimen with loop welded to valve floor, 4-3 mm in shell width (after Cooper 198 k);
B, Ecnomiosa inexpectata , with a bilacunar loop, 12 mm in shell width (after Cooper 198 k); c, Ecnomiosa
gerda , showing an adult loop connected to the median septum by a pair of medio-vertical connecting bands,
27-6 mm in shell width (after Cooper 1977); d, Macandrevia cranium , with a loop at the bilacunar stage, 4-5 mm
in shell length (after Davidson 1886); E, Macandrevia americana , showing a teloform adult loop and peculiar
cardinalia with a pair of hinge plates joining the valve floor, 25 0 mm in shell length (after Cooper 1982);
F, Zeilleria leckenbyi, with a teloform loop, 12-2 mm in shell length (after Baker 1972). Note the development
of spiny projections from the loop (labelled s) in these three genera.

species. The loop development and other morphological characters of two other species, from the
Caribbean Sea and the Indian Ocean, have been illustrated and described by Cooper (1977, 198k/).

The adult loop pattern of Ecnomiosa is characterized by the presence of medio-vertical connecting
bands and the absence of the lateral connecting bands (Text-fig. 9c; loop terminology as used by
Richardson 1975). This is certainly an advanced phase of the bilacunar phase of the Laqueidae sensu
Richardson ( 1975), as the loop of a younger stage retains the lateral connecting bands (the bilacunar
phase; Text-fig. 9b; see Cooper, 1977, pi. 35, figs 13-14; 198k/, pi. 14, figs 7 II). Richardson ( 1975)
did not list a genus with the adult loop pattern of Ecnomiosa , but listed the genera with the adult
loop of the bilacunar pattern as Aldingia , Jolonica , Kingena , and Paraldingia. Of these, Aldingia and
Pciraldingia have strikingly similar internal and external morphologies to that of Ecnomiosa , such
as shell outline, prominent concentric growth lines, tendencies to sulcation, cardinalia organization,
and, of course, the loop pattern. Indeed, the only major morphological features distinguishing the
former two genera from Ecnomiosa appears to be the presence or absence of the lateral connecting
bands and the extent of calcifications in the cardinalia and the loop. Aldingia had been assigned in
the Kraussinidae or the subfamily which used to embrace the kraussinid genera (Thomson 1927;
Muir-Wood et al. 1965), until Richardson (1975) reassigned it to the Laqueidae. Paraldingia was
erected by Richardson (1973) for the species which differ from Aldingia in possessing the dental

368

PALAEONTOLOGY, VOLUME 37

plates, more excavated and thinner cardinalia, and the thicker, broader, and spinous loop elements.
The presence of the dental plates suggests that the affinity of Ecnomiosa lies more with Paraldingia
than with Aldingia.

The prominent spinosity of the ascending and descending branches of the juvenile loop is another
major feature of Ecnomiosa morphology (Text-fig. 9b; see Cooper 1981c/, pi. 14, figs 7-14, 16-20).
This is the very feature that is also observed in Macandrevia and almost completely lacking, except
for the only slightly-developed spinous outgrowths from the septum (Richardson, 1975), in the
species of the other long-looped terebratulide group (terebratellids, Taqueids’, and dallinids)
recognized in this study. Although Elliott (1953) noted the spiny nature of many juvenile and adult
dallinid loops, an observation which was cited by Atkins (1959) in the study of Macandrevia , it may
be noteworthy that Macandrevia was then thought to be one of the most typical dallinids, and as
far as the dallinid species (as then considered) used in this study except for Macandrevia are
concerned (e.g. Dallina and Dallinella figured in Thomson 1927), the spinous outgrowths are
distinguishably trivial compared with those of Macandrevia or Ecnomiosa.

Another unusual juvenile feature of Ecnomiosa which warrants mention is the welding of the
descending lamellae to the valve floor during the early growth stages when only descending branches
and the septal pillar are developed (Text-fig. 9a; Cooper 1981c/, pi. 14, fig. 15). The meeting of the
descending branches to the valve floor is also observed in the adults of Argyrotheca , Megathiris , and
Gwynia (all megathyridids), and a Jurassic genus, Zellania , whose familial allocation in the
Terebratelloidea is uncertain (Muir-Wood et al. 1965). No welding is observed in any other
terebratulide species included in this study.

The loop development of Macandrevia is well documented (Friele 1877; Davidson 1886; Elliott
1953; Richardson 1975), and the basic features are certainly those of the typical long-looped species
except for the spinous projections (Text-fig. 9d), and the very early abortion, for a dallinid species,
of the connection between the loop and the septum (as noted by Smirnova 1984). The cardinalia,
on the other hand, are unique, with the inner hinge plates sloping to join the valve floor separately
(Text-fig. 9e). This feature is considered to reflect the unusual organization of the muscles of
Macandrevia , in particular the dorsal adjustors (Cooper 19736).

The mode of attachment of the dorsal adjustors to the dorsal valve could therefore be considered
as a common, but not exclusive, motif for the Cc//o/?//av3-M//ca///7/xu'//7-kraussinids-megathyridids
cluster, where the dorsal adjustor attachment sites appear to be confined to either the inner hinge
plates, or the valve floor (or both for Macandrevia). By contrast, in the species of the terebratellid-
L laqueid '-dallinid cluster, the dorsal adjustors never attach to the valve floor, but attach to parts of
the cardinalia, notably the outer and inner hinge plates. The development of the outer hinge plates
is known in Ecnomiosa , but they are narrow as described in Cooper (1977, p. 129), and their
function as sites of muscle attachment is doubtful.

From the discussions above, it may be concluded that the two long-looped genera, Ecnomiosa and
Macandrevia constitute a coherent group, and species of the other two families, the kraussinids and
megathyridids, may well have been derived by a process of neoteny from this long-looped group,
possibly independent of each other. This intepretation, however, is still inadequate in the light of
the fossil record and the affinity of this group to the Terebratuloidea.

Megathyridids and kraussinids are known from the Upper Cretaceous and Miocene, respectively
(Muir-Wood et al. 1965), while Macandrevia is known from the Miocene (Muir-Wood el al. 1965),
and no fossil is known for Ecnomiosa. From the distribution on both sides of the Atlantic and both
sides of the American continents in the present seas, Thomson (1927) suggested that the geological
history of Macandrevia should extend back at least to the Eocene. Fiving Macandrevia cranium is
known to occur in shallow cold waters, but most species, including M. cranium , range down to
1000-4000 metres (Cooper 1975). This abyssal distribution might explain the lack of a fossil record
for this genus. The geographical distribution of Ecnomiosa also suggests a long history of this genus.
Assuming a close relationship between Ecnomiosa , Aldingia , and Paraldingia , the record of this
group extends back to the Upper Eocene, since Aldingia and Paraldingia are known from the Upper
Eocene to Recent and the Upper Eocene to the Lower Miocene respectively (Richardson 1973).

ENDO ET A L TEREBRATULIDE BRACHIOPOD PHYLOGENY

369

New interpretation of ter ebratulide phytogeny

From the pattern and the spinous nature of the loop, it is possible to link Ecnomiosa with Kingena ,
and by analogy, the Kingenidae, which is known from the Upper Jurassic to Cretaceous (Owen
1970), or to Recent in the view of Smirnova (1984) who included Aldingia and Paraldingia in the
Kingenidae. Furthermore, from the spinous loop and the supposed general similarity in the way the
dorsal adjustor muscles attach to the dorsal valve, it may be possible to speculate an association of
these genera with the Zeilleriidae (cf. Text-fig. 9f). Morphological similarity between the
Terebratelloidea and Zeilleriidae was suggested by Richardson (1975), and further specifically
addressed by Elliott (1976). Baker (1972) demonstrated that the Jurassic Zeil/eria leckenbyi , a
‘typical’ zeilleriid species, involved the septal pillar in its loop development, a character often
attributed to the Terebratelloidea, and he insisted that the possession of spinose ascending and
descending elements is of prime importance for the inference of the zeilleriid ancestry.

Indeed, this pattern appears to be the best palaeontological solution to the puzzling disagreements
between the immunological and traditional schemes. The suggestion, therefore, is that the second
long-looped lineage recognized in the immunological study (K group) was derived from the
Terebratuloidea through the Zeilleriidae in Triassic times, when the Terebratuloidea and Zeilleriidae
started to be established and had morphological plasticity. The fact that the Zeilleriidae includes
two lineages, those which involve a septal pillar in the loop development, and those in which the
loop development takes place without the pillar (Baker 1972; Elliott 1976), ties in well with the
hypothesis of linking the long-looped K group species, which involve the septum in the loop
development, and the short-looped terebratuloids (which do not). Morphological similarity
between zeilleruds and Kingena has been noted by Muir-Wood et al. (1965). Smirnova (1984)
pointed out that Macandrevia loses the connection between the loop and septum very early during
ontogeny, unlike other ‘dallinoids’, a character which is suggestive of zeilleriid loop development.
Baker (1972, p. 470) noted the ‘resemblance between the early ascending lamellae of Z. leckenbyi
and the two divergent plates which constitute the early development of the loop of Kraussina ’, an
observation which fits perfectly with the interpretation that Kraussina originated from one of the
descendants of the zeilleriids. Baker (1972) also reported the similarity of the juvenile loop pattern
of Zeilleria leckenbyi with the adult loop pattern of Kingena , Zittelina (Kingenidae; Owen 1970),
and Trigonellina (Dallinidae; Muir- Wood et al. 1965), although the possibility of neotenous
evolution was denied on the basis of the presence or absence of the dental plate. Dagys (1972) and
Smirnova (1984) also considered dental plates to be of high taxonomic value in terebratulide
classification, and included kraussinids in the superfamily Terebratelloidea which generally lacks
dental plates, and considered that this superfamily is distantly related to the other extant long-
looped superfamily Dallinoidea, or to the extinct Zeilleriidae, both of which generally possess dental
plates.

The dental plates appear to constitute a fairly invariable character at family level or even in higher
classification. Although undoubtedly important, this character may not always be reliable at the
higher classifications. Among the dallinids, juvenile Dallina has dental plates, but the adults have
none, while in Campages , they are never present (Muir-Wood et al. 1965). The closely related
genera, Aldingia and Paraldingia , can be distinguished by the presence or absence of the dental
plates (Richardson 1973). Dental plates buttress teeth, so they can be interpreted as functioning to
strengthen the articulation between the valves. These structures may be advantageous in nonstrophic
brachiopods, such as most terebratulides, but in certain terebratulide taxa with a strophic tendency,
such as megathyridids and kraussinids. dental plates may well have degenerated secondarily by the
loss of functional constraints, since the posterior edges of the shell, being more or less parallel to
the hinge line, would provide some support for articulation. These morphological changes could
well have occurred independently in these taxa.

Cooper ( 1983, p. 51 ) noted that the early Triassic genus Vex had a 'loop, which suggests that of
Zeilleria but assigned it to the Terebratuloidea, rather than the Zeillerioidea, because of the lack
of median septum and dental plates. Among the early Triassic terebratulides. Cooper ( 1983) also
noted that Portneufia , first described as a short-looped dielasmatid genus (Hoover 1979), had a

370

PALAEONTOLOGY, VOLUME 37

text-fig. 10. A possible sequence of terebratulide phylogcny. See text for explanation. Brachiopod figures after
Cooper (1983), Baker (1972), and Muir-Wood et al. (1965).

similar cardinalia morphology to Vex, in addition to a long loop, but discriminated it from Vex by
the presence of dental plates (Cooper 1983). It is possible that these genera gave rise to the later
terebratuloids and zei 1 leriids respectively, the latter further producing kingenids and other K group
species, establishing the support for the loop from the median septum (Text-fig. 10).

Under this new interpretation based on immunological data, the radiation of the three major
lineages recognized in this study, terebratuloids, cancellothyridoids, and 1 terebratelloids must

ENDO ET AL. . TEREBRATULIDE BRACHIOPOD PHYLOGENY

371

have occurred prior to the divergence between terebratuloids and zeil leriids in the Triassic. In the
Palaeozoic, both the short-looped and long-looped lineages existed after early Devonian times
(Muir-Wood el al. 1965). Therefore the time of radiation for the last common ancestors of existing
terebratulides may be most parsimoniously assumed as the early Devonian (Text-fig. 8d), although
questions about Palaeozoic ancestors of the three major lineages remained open. Among the most
obscure is the origin of cancellothyridoids, as their ancestor is expected to have radiated at a similar
time in the Devonian from the other two major lineages, but direct evidence is wanting, since the
widely accepted earliest record of the cancellothyridoids is the Upper Jurassic (Muir-Wood et al.
1965). Taking Pseudokingena into consideration, which Cooper (1983) suggested as a member of
cancellothyridoids, the earliest record would still only be in the Lower Jurassic. From the
serotaxonomic viewpoint, cancellothyridoids are clearly separated not only from terebratelloids but
also from terebratuloids, a fact which appears to be supported by the unusual morphology of the
cancellothyridoids (cf. Cooper 1973a). Clearly, however, their origin remains enigmatic.

CONCLUSIONS

The serotaxonomic data suggest radical new ideas about the higher-order phylogeny within the
Terebratulida. The immunological approach was also applied with reasonable success to the
inference of lower-order relationships, down to species-level within the genus Terebratulina , using
the more elaborate inhibition method. The serotaxonomic data have major implications for
terebratulide classification, especially at subordinal and superfamilial arrangements. New
taxonomic units and rearrangements are required for these ranks, but we are withholding the
presentation of formal nomenclatorial recommendations, because of the absence in our study, due
to the difficulty in obtaining specimens, of such important living taxa as Platidiidae, Thaumatosiidae,
Phaneroporiidae, and Magadinae.

The immunological view does not flatly contradict traditional taxonomy, but urges further
morphological investigations of the adult and juvenile stages of both Recent and fossil taxa, in
particular those that lived during the late Palaeozoic and Mesozoic times. It is possible that the
difficulty in observing loop development of fossil species, a procedure which requires a sample with
a good ontogenetic sequence and often the use of serial sectioning, has been hindering the
recognition of certain affinities among fossil taxa, especially of features which are only observed in
juveniles. The apparent contradictions between the morphological and molecular data are only
problematical if one or the other source of information is regarded as rigid and definitive; this would
be entirely unrealistic, for brachiopods and all other organisms. The inferences from molecular data
must never be assessed in isolation, but must instead be cross-checked with both geological
information, morphology, and other molecular techniques, of which the sequencing of nucleic acids
would be the logical next target in the hope of resolving some of the major classification problems
in brachiopods.

Acknowledgements. We thank Drs C. H. C. Brunton, R. E. Grant, A. Ansell, C. Emig, M. James, L. Peck, and
Mr E. Tsuchida for kindly providing us with brachiopod samples for this study. Thanks are due to Dr S. Ewing
for preparing some brachiopod antisera. Dr D. MacKinnon is thanked for constructive criticisms and for
translations of Russian literature. The research was supported by grants from the Systematics Association
(K.E.), Shimadzu Foundation, Kyoto (K.E.), Royal Society (M.J.C. and G.B.C.), NERC (M.J.C. and
G.B.C.). the Lounsbery Foundation, New York (G.M. and P. W.), and NSF (P. W.).

REFERENCES

atkins, D. 1959. The early growth stages and adult structure of the lophophore of Macandrevia cranium
(Muller) (Brachiopoda, Dallinidae). Journal of the Marine Biological Association , 38, 335-350.
baker, p. g. 1972. The development of the loop in the Jurassic brachiopod Zeilleria leckenbyi. Palaeontology ,
15, 450-472.

372

PALAEONTOLOGY, VOLUME 37

brunton, c. h. c. and hiller. n. 1990. Late Cainozoic brachiopods from the coast of Namaqualand, South
Africa. Palaeontology , 33, 313-342.

cohen, b. l. 1992. Utility of molecular phylogenetic methods: a critique of immuno-taxonomy. Lethaia, 25,
441^142.

COLLINS, M. J., CURRY, G. B., MUYZER, G., QUINN, R., XU, S., WESTBROEK, P. and EWING, S. 1991fl. Immunological
investigations of relationships within the Terebratulid brachiopods. Palaeontology , 34, 785-796.

— westbroek, p., zoMERDHK, t. and quinn, r 1988. Serotaxonomy of skeletal macromolecules
in living terebratulid brachiopods. Historical Biology , 1, 207-224.

muyzer, G., curry, g. b., Sandberg, p. and westbroek, p. 1991/7. Macromolecules in brachiopod shells:
characterization and diagenesis. Lethaia, 24, 387-397.

cooper, g. a. 1973«. Fossil and Recent Cancellothyridacea (Brachiopoda). Tohoku University Science Reports,
2nd Series (Geology) ( Special Volume), 6. 371-390.

1973/7. Verna's Brachiopoda (Recent). Smithsonian Contributions to Paleobiology, 17, 1-51.

1975. Brachiopods from West African waters with examples of collateral evolution. Journal of
Paleontology, 49, 91 1-927.

1977. Brachiopods from the Caribbean Sea and adjacent waters. University of Miami Press, Florida,

21 1 pp.

1979. Tertiary and Cretaceous brachiopods from Cuba and the Caribbean. Smithsonian Contributions to
Paleobiology, 37, 1-45.

1 98 1 z/. Brachiopoda from the southern Indian Ocean (Recent). Smithsonian Contributions Jo Paleobiology ,
43. 1-93.

I9816. Brachiopoda from the Gulf of Gascogne, France (Recent). Smithsonian Contributions to
Paleobiology , 44, 1-35.

— 1982. New brachiopods from the southern hemisphere and Cryptopora from Oregon (Recent).
Smithsonian Contributions to Paleobiology, 41, 1—43.

1983. The Terebratulacea (Brachiopoda), Triassic to Recent: a study of the brachidia (loops).
Smithsonian Contributions to Paleobiology 50, 1-445.

— 1989. Jurassic brachiopods of Saudi Arabia. Smithsonian Contributions to Paleobiology , 65, 1-213.

— and endo. k. 1991 . Migration of brachiopod species in the N. Atlantic in response to Pleistocene climate
change. Geology , 19. 1101-1103.

-QUINN, R., COLLINS, M. J., ENDO, K„ EWING, S., MUYZER, G. and WESTBROEK, P. 1 99 1 <7. Immunological
responses from brachiopod skeletal macromolecules; a new technique for assessing taxonomic relationships
using shells. Lethaia, 24, 399-407.

— walton, d., endo, K., clegg, h., Abbott, G. and Armstrong, h. 1991/7. Biogeochemistry of
brachiopod intracrystalline molecules. Philosophical Transactions of the Royal Society of London, Series B,
333, 359-366.

curry, G. b., cusack, m., endo, k., walton, d. and quinn, r. 1991c. Intracrystalline molecules from
brachiopods. 35-39. In SUGA, s. and nakahara, h. (eds). Mechanisms and phytogeny of mineralization in
biological systems. Springer-Verlag, Tokyo, 500 pp.

dagys, a. s. 1968. [Jurassic and early Cretaceous brachiopods of North Siberia]. Akademiya Nauk SSSR.
Sibirskoe Otdelenie , Trudy Instituta Geo/ogii i Geofiziki. 41, 1-160. [In Russian].

1972. [The postembryonic development of the brachidium of late Paleozoic and early Mesozoic
terebratulids]. Akademiya Nauk SSSR, Sibirskoe Otdelenie , Trudy Instituta Geologii i Geofiziki, 112, 21 -58.
[In Russian].

davidson, T. 1886. A monograph of Recent Brachiopoda. Part I. Transactions of the Linnean Society of
London, 2nd Series, 4, 1-74.

elliott, g. f. 1953. Brachial development and evolution in terebratelloid brachiopods. Biological Reviews, 28,
261-279.

1976. Comments on ‘The loop development and the classification of terebratellacean brachiopods'.
Palaeontology, 19, 413-414.

friele, h. 1877. The development of the skeleton in the genus Waldheimia. Archiv for Mathematik og
Naturvidenskab , 2, 380-386.

hoover, p. r. 1979. Early Triassic terebratulid brachiopods from the western interior of the United States.

United States Geological Survey Professional Paper , 1057, 1-21.
johnstone, a. and thorpe, R. 1987. Immunochemistry in practice. (2nd edition). Blackwell Scientific
Publications, Oxford, 306 pp.

ENDO ET AL.: TEREBRATULIDE BRACHIOPOD PHYLOGENY

373

lesperance, p. j. 1990. Cluster analysis of previously described communities from the Ludlow of the Welsh
Borderland. Palaeontology , 33. 209-224.

muir-wood, h. M., stehli, f. g., elliott, g. f. and hatai, R. 1965. Terebratulida. H728-H857. In moore,
r. c. (ed.). Treatise on invertebrate paleontology. Part H. Brachiopoda. Geological Society of America
and University of Kansas Press, Boulder, Colorado and Lawrence, Kansas. 927 pp.

owen, E. F. 1970. A revision of the brachiopod subfamily Kingeninae Elliott. Bulletin of the British Museum
of Natural History , Geology Series , 19. 27-83.

richardson, j. r. 1973. Studies on Australian brachiopods. 2. The family Laqueidae (Terebratellidae).
Proceedings of the Royal Society of Victoria , 86, 1 17-125.

1975. Loop development and the classification of terebratellacean brachiopods. Palaeontology. 18.
285-314.

ride, w. d. l., sabrosky, c. w., bernardi, G. and melville, R. v. (eds). 1985. International code of zoological
nomenclature. (3rd edition). University of California Press, London, Berkeley and Los Angeles, 338 pp.

Smirnova, T. N. 1984. [Lower Cretaceous brachiopods (morphology, systematics, phytogeny and their significance
in biostratigraphy and palaeozoogeography)\. Nauka Press, Moscow, 199 pp. [In Russian],

sneath, p. h. a. and sokal, R. R. 1973. Numerical taxonomy. W. H. Freeman, San Francisco, 573 pp.

Thomson, j. a. 1927. Brachiopod morphology and genera (Recent and Tertiary). New Zealand Board of Science
and Art, Manual, 7, 1 338.

williams, a. and hurst, j. m. 1977. Brachiopod evolution. 79-121. In hallam, a. (ed.). Patterns of evolution
as illustrated by the fossil record. Elsevier, Amsterdam, 591 pp.

- and eighteen others 1965. Brachiopoda. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Part
H. Geological Society of America and University of Kansas Press, Boulder, Colorado and Lawrence,
Kansas, 927 pp.

K. ENDO1
G. B. CURRY

Department of Geology and Applied Geology
The University
Glasgow G12 8QQ, UK

R. QUINN

Department of Immunology
The University
Glasgow G1 1 6NT, UK

M. J. COLLINS

NRG, Drummond Building
The University

Newcastle Upon Tyne NE1 7RU, UK

G. MUYZER

MPI for Marine Microbiology
Molecular Ecology Unit
Fahrenheitstr. I
2800 Bremen 33, Germany

P. WESTBROEK

Geobiochemistry Unit, Gorlaeus Laboratory
Leiden University
Leiden, The Netherlands

1 Present address
Geological Institute

Typescript received 16 April 1993 University of Tokyo

Revised typescript received 2 December 1993 Tokyo 113, Japan

MIDDLE TO LATE TELYCH IAN (SILURIAN:
LLANDOVERY) GRAPTOLITE ASSEMBLAGES OF

CENTRAL WALES

by JAN ZALASIEWICZ

Abstract. Graptolite faunas from the uppermost turriculatus , crispus , griestoniensis , crenulata and spiralis
Biozones of central Wales are described. Detailed collecting integrated with recent BGS mapping has
demonstrated a new subzone of Torquigraptus carnicus in the uppermost turriculatus Biozone, a subdivision
of the crispus Biozone into three successive subzones of, in ascending order: Monoclimacisl galaensis ,
Monograptus crispus , and Streptograptus sartorius ; an informal subdivision of the griestoniensis Biozone into
two successive assemblages, and the presence of faunas which can be assigned to the late Telychian crenulata
and, probably, the spiralis Biozones. The compositions of these successive assemblages are outlined, and
selected graptolites are figured and described. These include two new species: Monograptus pseudocommunis
and Streptograptus whitei.

Most early studies of the Llandovery graptolite faunas from central Wales concentrated on
assemblages from Rhuddanian and Aeronian strata (e.g. Lapworth 1900; Jones 1909; Davies 1929;
Sudbury 1958) and a number of the standard British graptolite biozones stem from those works (see
Rickards 1976; Zalasiewicz 1990). Upper Llandovery (Telychian) faunas were, until recently,
relatively neglected, because of the great thickness and structural complexity of the Telychian
deposits of central Wales. Only one of the early major studies (Wood 1906 on the Tarannon area)
focused on this part of the sequence. More recent work has, however, demonstrated the
biostratigraphical potential of the well-preserved graptolite faunas preserved within parts of the
thick Telychian turbidite sequences of Wales. Loydell (1991, 1992 a, 19936) in particular, has
effected a detailed subdivision of the turriculatus Biozone in west-central Wales.

Recent mapping by the British Geological Survey of the Rhayader and Llanilar districts (Text-
fig. 1 ; Davies et al. in press) has enabled many sections yielding faunas referable to the uppermost
turriculatus , crispus , griestoniensis , and crenulata Biozones of the standard British sequence (as
outlined in Rickards 1976, 1989; Zalasiewicz 1990) to be examined. This paper summarizes the
biostratigraphical results of this BGS work. The sequence of faunas is outlined, and descriptions are
made of selected taxa, including two new species. All fossil material referred to is lodged with the
British Geological Survey, Keyworth, unless stated otherwise.

GEOLOGICAL SETTING

Details of the stratigraphy, and of the main sections studied, are given in Davies et al. (in press). The
succession comprises a number of thick turbidite formations. The hemipelagic intervals between
individual turbidites are commonly bioturbated but otherwise unfossiliferous. Some stratigraphical
intervals, though, contain hemipelagites in which a fine primary lamination, the result of deposition
in anoxic bottom conditions (Cave 1979) is preserved. Such hemipelagites contain a well-preserved
graptolite fauna as do, locally, some turbidite sandstones. The graptolite-bearing hemipelagites tend
to be most common in strata of late turriculatus and crispus Biozone age; they are rarer in the

| Palaeontology, Vol. 37, Part 2, 1994, pp. 375-396. |

© The Palaeontological Association

376

PALAEONTOLOGY, VOLUME 37

text-fig. 1. Location of study area
(boxed) in Wales. Outcrop of Llandovery
sediments shown by stippled ornament.

griestoniensis Biozone and scarce in younger Telychian strata. Continuously anoxic sea floor
conditions were re-established close to the Llandovery-Wenlock boundary (Davies et al. in press).

PRESERVATION

The Llandovery graptolites of the Rhayader and Llanilar district are mainly preserved in partial to
full relief. Fragmentary periderm is preserved frequently, and usually surrounds an internal mould
of pyrite or limonite. Specimens infilled by sediment or preserved as hollow external moulds are less
common. Although often showing well-preserved thecal structures, commonly the graptolites were
obtained only as partial rhabdosomes because, during collection, the rock fractured along cleavage
and not bedding. This caused particular problems with the recognition of some graptolite taxa,
notably the spiraliform monograptids. Graptolites in turbidite sandstones commonly lack a pyrite
infill and are mostly diagenetically flattened.

The graptolites are variably deformed, depending on the mode of preservation, degree of
tectonism and enclosing lithology. Brittle fracture of pyritized graptolites is typical, the graptolites
being broken into segments at their weakest points; the segments tend to be arranged in an
imbricate fashion where the long axis of the rhabdosome is perpendicular to the cleavage lineation.
Where possible, dimensions of deformed graptolites were taken parallel to this lineation; in this
orientation, the pyrite internal moulds tended to resist tectonic compression, while the nature of the
brittle fracture indicated that extension parallel to the long axis of the rhabdosome was commonly
slight.

BIOSTRATIGRAPHY

This paper summarizes the graptolite biostratigraphy from the upper part of the turriculatus
Biozone to the upper part of the crenulata Biozone sensu Rickards (1976, 1989), an interval here
referred to the spiralis Biozone by comparison with sequences in mainland Europe. The

ZALASIEWICZ : LATE LLANDOVERY GRAPTOLITES

377

biostratigraphy outlined here has been applied within the central Welsh Basin; it is not yet known
to what extent it is applicable elsewhere. Ranges of taxa reported from other sequences (e.g. Boucek
1953) differ in detail, and it is not yet clear whether this is due largely to faunal diachroneity or to
problems of identification. The ranges of taxa are given in Text-figure 2.

Spirograptus turriculatus Biozone

Loydell's (1991) subdivision of the turriculatus Biozone was employed during the survey, although often it was
not possible to distinguish between the lowest three subzones. One additional subzone, that of Torquigraptus
carnicus , was employed in the survey to characterize strata forming the uppermost part of the turriculatus
Biozone. A subsequent zonation by Loydell (1992u) separated a biozone of Spirograptus guerichi (comprising
the runcinatus to lower u tilis Subzones) from the lower part of the turriculatus Biozone. The relations of these
subzonal schemes is shown in Table 1.

Torquigraptus carnicus Subzone

This subzone is marked by the incoming of Torquigraptus carnicus (Gortani, 1923), Streptograptus whitei
sp. nov., S. exiguus (Lapworth, 1876) and Monoclimacisl galaensis (Lapworth, 1876). Species ranging up from
the proteus and earlier subzones include S. storchi Loydell, 1991, which is common, and the hooked
monograptids Monograptus marri Perner, 1897, and M. rickardsi Hutt, 1975. 5. whitei and S. storchi have not
yet been found in the overlying crispus Biozone though, in one instance, in the A44 road section near Llangurig
(National Grid Reference SN 8580 8198), S’. storchi has been found associated with M. aff. crispus , a narrower
and more densely thecate form than M. crispus Lapworth, 1876. In the Czech Republic, S. storchi has been
noted in the lower part of the crispus Biozone (D. K. Loydell, pers. comm.).

Monograptus crispus Biozone

This is marked by the incoming of M. crispus Lapworth, 1876, and M. discus Tornquist, 1883 (= M. veles
Richter, 1871). Neither is common near the base of the biozone and so recognition of the biozonal boundary
locally is difficult. It is not known with certainty which of the two species appears first in central Wales.
Bjerreskov (1975) showed the two species as appearing simultaneously on Bornholm, but in a graptolitic level
above barren beds. The more continuously fossiliferous sections in Germany (Schauer 1971) and the Czech
Republic (Boucek 1953) indicate that M. discus appears before M. crispus. Three successive assemblages can
be recognized in the crispus Biozone of central Wales, and these are here formalized as subzones. They are, in
ascending order:

1 . Monoclimacis ? galaensis Subzone

This is a partial-range subzone, its base defined by the overlap of Monoclimacisl galaensis , and T. carnicus
with M. crispus , and M. discus. Petalolitlius sp. I of this paper and Monograptus aff. crispus are also present.
Mcll galaensis , T. carnicus , Pe. sp. 1, and M. aff. crispus have not been found to range higher in central Wales.
M. clintonensis , and S. exiguus are common, but M. crispus and M. discus occur only sporadically. The
importance of Monoclimacis galaensis as an indicator of high turriculatus to low crispus Biozone strata has been
noted by Bjerreskov (1975) and Rickards (1976).

2. Monograptus crispus Subzone

This is a partial range subzone defined by the overlap of M. crispus and Streptograptus loydeUi Storch and
Serpagli, 1993. Both taxa are abundant.

3. Streptograptus sartorius Subzone

This is the interval between the disappearance of M. crispus and the first appearance of the monoclimacid
monograptids which characterize the griestoniensis Biozone. Torquigraptus pragensis pragensis (Pribyl, 1943),
T. pragensis ruzickail (Pribyl, 1943), and Streptograptus sartorius (Tornquist, 1881 ) appear for the first time.

378

PALAEONTOLOGY, VOLUME 37

Pristiograptus nudus s.l.

Spirograptus turriculatus X X

Monograptus bjerreskovae X ?

Normalograptus? nebula X X

Monograptus marri X cf.

Streptograptus storchi X A

Monograptus rickardsi X X

Petatotithus tenuis si X X

Streptograptus pseudobecki X

Torquigraptus proteus X X

Monoclimacis? galaensis ? X

Monograptus aff . rickardsi (slender form) X

Torquigraptus carnicus X

Streptograptus whitei X

Monograptus aff. crispus R

Streptograptus exiguus X

Monograptus clintonensis ?

Streptograptus mustadi Howe 1982 ms
Monograptus crispus
Monograptus discus
Petatotithus sp. 1
Streptograptus loydelli
Retiotites geinitzianus subspp.

Torquigraptus pragensis ruzickai ?

Torquigraptus pragensis pragensis
Streptograptus aff. sartorius
Streptograptus sartorius
Torquigraptus tullbergi spiraioides
Monoclimacis cf. griestoniensis of Elies and Wood
Monoclimacis griestoniensis s.s.

Torquigraptus tullbergi tullbergi
Streptograptus aff. loydelli
Monograptus priodon
Monograptus pseudocommunis
Torquigraptus pergracilis?

Monograptus sp. 1
Pristiograptus initialis

Torquigraptus ex. gr. pragensis ? (slender form)
Monoclimacis vomerina vomerina
Monoclimacis cf. crenulata sensu Elies and Wood
Monoclimacis linnarssoni
Monoclimacis griestoniensis nicoli
Monograptus spiralis
Monograptus aff. falx
Monograptus parapriodon
Pristiograptus? aff. initialis (broad form)

?

cf.

X
cf .

X
cf .

text-fig. 2. Range-chart of species from the latest turriculatus Biozone (carnicus Subzone) to the spiialis
Biozone in the area studied. X, taxon present; A, taxon abundant; R, taxon rare.

ZALASIEWICZ : LATE LLANDOVERY GR APTOLITES

379

table I Subdivisions of the turriculatus Biozone used in west and central Wales.

Loydell (1991)

Loydell (1992a)

Davies et al. (1994)

crispus Biozone

crispus Biozone

turriculatus

Biozone

' (?)

proteus Subzone
johnsonae Subzone
utilis Subzone
renaudi Subzonc
gemmatus Subzone
„ runcinatus Subzone

turriculatus Biozone
guerichi Biozone

carnicus Subzone \

proteus Subzone

johnsonae Subzone

utilis Subzone . ,

c , turriculatus Biozone

renaudi Subzone

gemmatus Subzone

runcinatus Subzone

(not recognized) t

Many assemblages in this subzone, as in the upper levels of the preceding turriculatus Biozone, are dominated
by relatively non-diagnostic monograptids resembling M. marri and M. clintonensis (sensu Hall, 1852). A
comparable interval, lacking both M. crispus and monoclimacids, was recognized by Wilson (1954) in the
Howgill Fells, northern England and by Bjerreskov (1975) on Bornholm.

griestoniensis Biozone

The base of this biozone is recognized by the appearance of two slender monoclimacids: Mcl. griestoniensis
(Nicol, 1850), and the distinctive Mcl. cf. griestoniensis of Elies and Wood (1911). The latter appears slightly
earlier in central Wales (see Zalasiewicz 1990, fig. 2; Davies et al. in press) and dominates the lower part of
the biozone, along with species extending up from the upper part of the crispus Biozone, such as T. pragensis
pragensis , S. sartorius , and T. tullbergi spiraloides (Pribyl, 1945). Mcl. griestoniensis , rare in the lower part of
the biozone, becomes common in the upper part.

crenulata Biozone

The crenulata Biozone, as traditionally recognized in Britain (e.g. Rickards 1976) occurs between the
griestoniensis Biozone and the centrifugus Biozone. In Europe several other biozones have been recognized
between a restricted crenulata Biozone and the centrifugus Biozone (e.g. Boucek 1953; Bjerreskov 1975). These
are difficult to recognize in Britain, partly because of a relative scarcity of anoxic, graptolite-bearing levels, and
partly because of a scarcity of diagnostic fossils around the Llandovery-Wenlock boundary. However,
elements of this more precise biostratigraphy have been recently identified in the UK (Loydell 1993a; Loydell
and Cave 1993), and thus the crenulata Biozone has here a correspondingly restricted definition.

The incoming of broad vomerinids, particularly Md. vomerina vomerina (Nicholson, 1872) is used to denote
the lower boundary of this biozone; the zonal fossil, Mcl. crenulata (sensu Elies and Wood 1911) has been
recognized at one locality (SO 0817 7417). Accompanying taxa include species continuing from earlier biozones
such as M. discus, and Torquigraptus tullbergi tullbergi (Boucek 1931).

Monograptus spiralis Biozone

A single locality (SO 0092 7707) recorded in the highest graptolitic unit in the Dolgau Mudstones (the
uppermost Telychian formation of the Rhayader/Llanilar districts; Davies et at. in press), yielded newcomer
species such as Monoclimacis linnarssoni (Tullberg, 1883), Monograptus parapriodon Boucek, 1931, and M. aff.

380

PALAEONTOLOGY, VOLUME 37

ZALASIEWICZ : LATE LLANDOVERY GRAPTOLITES

381

falx (Suess, 1851), while M. discus and T. tullbergi subspecies are absent. This assemblage seems to correlate
with the upper part of the spiralis Biozone of Bjerreskov (1975), and with the spiralis Biozone of Boucek ( 1953).
Monograptus spiralis (Geinitz, 1842) has been found only rarely elsewhere, occurring in isolated localities
without any diagnostic accompanying fauna.

PALAEONTOLOGICAL NOTES

Petalolitluis sp. 1 of this paper (Text-fig. 3e-h) is a smaller and more densely thecate species than
comparable petalograptids such as P. tenuis , and related diplograptids such as the Orthograptus sp.
of Hutt et al. (1970). It most closely resembles Petalolithus kurcki (Rickards, 1970; see Text-fig. 3i
and redescription in Loydell 1992a) from the sedgwickii to lower turriculatus Biozones, but differs
in having a more convexly curved ventral wall of theca l* 1.

Monoclimacis cf. griestoniensis of Elies and Wood (1911) (Text-fig. 6q-t) is narrower than
Monoclimacis crenulata ( sensu Elies and Wood, 1911) (Text-fig. 8h-k; see Loydell et al. 1992 for a
discussion of the nomenclature of this species) and more robust than Monoclimacis griestoniensis
(Text-fig. 6u), with a different proximal and thecal morphology.

The material assigned tentatively to Monoclimacis griestoniensis nicoli Rickards, 1965 (Text-fig.
8l-m) resembles the type description. But, except for its narrowness, it also resembles Mcl.
linnarssoni found at the same horizon; closer study may show that it lies within the latter’s range
of variability. The material here assigned to linnarssoni (Text-fig. 8n-q) agrees well with the detailed
description given by Bjerreskov (1975, p. 55), the only difference being that Bjerreskov’s material
has a longer sicula (c. L8 mm) and theca 1. Rickards’ (1965) material assigned to Mcl. linnarssoni
from the centrifugus Biozone of the Howgill Fells appears not to be conspecific. The rhabdosome
in that material is more rapidly expanding, reaching 0-7 mm at thlO, has a maximum width of
1-7 mm, and the supragenicular walls appear quite straight rather than possessing a distinct convex
curvature.

Streptograptus storchi (Text-fig. 3n) is a more robust ‘straight streptograptid ’ than the
approximately contemporaneous species S. pseudobecki (Boucek and Pribyl, 1942; Text-fig. 3m),
which has been redescribed by Loydell (1990a).

Streptograptus loydelli Storch and Serpagh (Text-fig. 4i-l) can be distinguished from S. exiguus
(Text-fig. 3d) by its greater width, the more marked distal expansion of the prothecae and by the

text-fig. 3. Graptolites from the turriculatus and lower crispus Biozones, a-c, Torquigraptus carnicus (Gortani,
1923); all on block BGS JZ1034; turriculatus Biozone, carnicus Subzone; quarry (SN 94457650) E of
Cwmgwary Farm, 8 km NNW of Rhayader, d, Streptograptus exiguus (Nicholson, 1868); BGS JZ2173,
turriculatus Biozone, carnicus Subzone; track section (SN 81876035) Twyi Forest, e-h, Petalolithus sp. I ;
crispus Biozone, probably galaensis Subzone; stream section (SN 95875884) SW of Newbridge-on-Wye;
e. BGS JZB39; F, BGS JZB40 (counterpart of e); g, BGS JZB35; h, BGS JZB33 (counterpart of G).

i, Petalolithus kurcki (Rickards, 1970); BGS JZ6160; turriculatus Biozone, Irenaudi Subzone; quarry near
Bwlch-y-ddault, Dyfed (SN 71096309). j-l, Monograptus aff. crispus Lapworth, 1876; BGS DJ8292; probably
turriculatus Biozone, carnicus Subzone; A44 near Llangurig (SN 85808198). K., BGS JZ984; lower crispus
Biozone; Cwmgwary Farm (SN 94457650). m, Streptograptus pseudobecki (Boucek and Pribyl, 1942); BGS
JZ6961, turriculatus Biozone, proteus Subzone; cliff section (SN 55727494) 7 km SSW of Aberystwyth. N,
Streptograptus storchi Loydell, 19936; BGS DJ8297 ; turriculatus Biozone, carnicus Subzone; road cutting (SN
85809198) near Llangurig. o-Q, Streptograptus whitei sp. nov.; all turriculatus Biozone, carnicus Subzone; o,
BGS JZ1035; quarry (SN 94457650), 8 km NW of Rhayader; part of near-proximal fragment, p-q, DJ7612
(holotype), section (SN 85086545) by Claerwen Reservoir; mesial and distal parts.

All figures x 10 (larger scale bar represents 2 mm) except a, c, x 5 (smaller scale bar represents 2 mm).

382

PALAEONTOLOGY, VOLUME 37

text-fig. 4. For legend see opposite.

ZALASIEWICZ : LATE LLANDOVERY GRAPTOLITES

383

higher profile of the metathecae. S. plumosus (Baily, 1871), usually attains greater widths (of up to
0-9 mm) and has more robust prothecae (see Loydell 1990A). Streptograptus afif. loydelli (Text-fig.
6j-o) has the same overall dimensions, but differs in being spinose and in having a narrower theca
2. It may be related to ‘ Monograptus exiguus C' of Bjerreskov (1975), which is also spinose, from
the upper crispus Biozone of Bornholm. However, the single fragment of that taxon described is
slightly narrower (0-5 mm) and has parallel prothecal walls and slightly lower profile metathecae.

The material of Streptograptus sartorius (Text-fig. 5a-f) compares well with topotype material
examined (that figured by Tornquist in 1892), with the only observable difference being slightly
more widely spaced thecae in that material, with 2TRDs (2TRD = two theca repeat distance: Howe
1983) of 2-0-2- 1 mm. Some other topotype fragments, however, show widths of up to 045 mm in
relief, indicating more variation in the Swedish material than has been observed in that from central
Wales. The material figured by Boucek and Pribyl (1951) from the crenulata Biozone of Bohemia
appears different, with possibly hooked metathecae and more distinctly expanding prothecae.
Several rhabdosomes from one locality are here referred to S. aff. sartorius (Text-fig. 5g-k). These
differ only in possessing lateral spines, which possibly reflect enhanced preservation of a feature
present in all material of S. sartorius. S. sartorius can be confused with proximal fragments of
Monograptus crispus ; in that species, though, the proximal thecae are considerably more widely
spaced (see below).

Pristiograptusl aft', initialis (Kirste, 1919) (Text-fig. 8c-d) differs from the typical form (Text-fig. 8g;
= Monoclimacis shottoni Rickards, 1965) in its greater breadth and increased rate of expansion.

Torquigraptus carnicus (Text-fig. 3a-c) resembles a very slender form of T. proteus, with an even
more irregular helical spiral than that species possesses. Most fragments are 0-3-04 mm wide (max.
0-6 mm) and 2TRDs are c. 3-0 mm. Thecae appear laterally twisted throughout. It may be
conspecific with the material described as Monograptus angustus Rickards, 1970 by Strachan (1982,
p. 163, fig. 2h-i).

The combination of rastritiform thecae proximally and markedly asymmetrical thecae distally
renders Torquigraptus pragensis pragensis (Text-fig. 5l-p) a distinctive species. Previous descriptions
of this species (e.g. Storch and Serpagli 1993) have been based on flattened material and so have
not shown the lateral twisting of the metathecae. The pragensis group does not therefore comprise
the latest representatives of the triangulate monograptids (cf. Rickards 1989), but belongs to the
planus-proteus-tullbergi lineage of monograptids recognized by Bjerreskov (1975), i.e. to
Torquigraptus.

Torquigraptus pragensis ruzickail (Text-fig. 5q) differs from T. pragensis pragensis in having lower,
and fewer, rastritiform thecae proximally, which are inclined at a lower angle to the rhabdosome;

text-fig. 4. Graptolites from the crispus Biozone. A-H, Monograptus crispus Lapworth, 1876. a, BGS DJ7270;
b, enlargement of a, to show details of thecal structure, c, BGS SPT1977; sicula and proximal thecae, d, BGS
DJ7238; mesial theca, e, BGS DJ7268; mesial theca, f, BGS DJ7281 ; distal theca. G, BGS DJ728I ; distal
theca, h, BGS DJ7268; distal theca; a-b, d-h, crispus Biozone, crispus Subzone; north bank of Afon Ystwyth
(SN 83887547). c, crispus Biozone, crispus Subzone, Afon Claerwen (SN 82176929). i-l, Streptograptus loydelli
Storch and Serpagli, 1993; I— j, l, BGS DJ7223. k, BGS DJ7228; all crispus Biozone, crispus Subzone; Afon
Ystwyth (SN 83937548); line denotes cleavage trace, m-o, Torquigraptus aff. pergracilis ? (Boucek, 1931).
M, BGS SPT1916. N-o, BGS SPT1919; crispus Biozone, galaensis or crispus Subzone; stream section (SN
82016518), 2-15 km S of Claerwen, which is 15 km W of Rhayader, Powys, p-q, Monograptus c/intonensis
(sensu Hall, 1852); p, BGS DJ8370; crispus Biozone, trackside quarry (SN 89437820) W of Llangurig. Q, BGS
DJ9343; ? crispus Biozone; stream section (SN 85688020) W of Llangurig.
a, m-n, p-q, x 10 (smaller scale bar represents 2 mm); b-h, o, x 20 (larger scale bar represents 2 mm).

PALAEONTOLOGY, VOLUME 37

text-fig. 5. For legend see opposite.

ZALASIEWICZ: LATE LLANDOVERY GRAPTOLITES

385

it also shows a tendency towards gracilization of the extreme proximal end. In these respects it seems
morphologically intermediate between T. pragensis pragensis and torquigraptids possessing a gracile
proximal end such as T. proteus and T. tullbergi. The Monograptus pragensis ruzickai described by
Hutt ( 1975), a sedgwickii Biozone taxon, appears to differ in possessing high triangulate thecae that
are apparently symmetrical throughout the rhabdosome.

Torquigraptus ex. gr. pragensis ? (Text-fig. 6p) is tentatively interpreted as a member of the pragensis
group, in view of the near-straightness of the fragments and the relative breadth of the common
canal. It is narrower distally than T. pragensis pragensis.

Material of Torquigraptus tullbergi tullbergi (Text-fig. 8e-f) differs slightly from the holotype by a
more gentle dorsal curve and a slower rate of expansion proximally (width at theca 5 is 035 mm
and at theca 10 is 0 75 mm ; corresponding values for the tullbergi tullbergi holotype are 0-5 mm and
0-8 mm - P. Storch, pers. comm.). This species may be distinguished from other spirally curved
torquigraptids by slender, triangular prothecae and a gentle dorsal curvature. T. tullbergi spiraloides
(Text-fig. 7j-m) expands more rapidly (width at theca 5 is 0-75-0-9 mm) and has a more coiled
rhabdosome; examples of this taxon have been identified as M. spiralis in the past (e.g. Text-fig.
7k-m, from Grieston Quarry, Scotland), but the more slender proximal end of spiraloides serves as
a discriminating feature.

Monograptus clintonensis (Text-figs 4p-q, 7n) is a slender, slowly expanding monograptid the
hooked thecae of which bear lateral apertural spines which are most noticeable proximally. In the
past, it has been confused with Monograptus priodon (Text-fig. 8r; see Loydell 1992 6, for full
discussion), which expands more rapidly and is more densely thecate proximally. In general
dimensions, the species resembles Monograptus nrarri , which possesses shorter lateral horns, which
are not visible in most preservational modes (Hutt et al. 1970; Hutt 1975).

The material placed in affinity with M. falx (Text-fig. 8a- b) resembles M. spiralis in its thecal
structure and robust proximal end. However, it appears to be a distinctly shorter form, dorsally
curved rather than spirally coiled with a lower rate of expansion, more closely spaced thecae and
a maximum observed width (TO mm) that is considerably less than the 3 0 mm quoted for
M. spiralis s.s. It differs from M. falx in attaining slightly greater lengths than the 12 13 thecae
quoted as typical (Pfibyl 1945), and in possessing wider prothecal bases (D. K. Loydell, pers.
comm.).

The graptolite here termed Monograptus sp. I, is spiraliform, the single specimen recovered
comprising an incomplete whorl 65 mm in diameter. The thecae (Text-fig. 7o-p) are asymmetrical
but not laterally twisted, the right side of the metatheca growing farther than the left, so that the

text-fig. 5. Graptolites from the upper crispus Biozone ( sartorius Subzone) to the lower griestoniensis Biozone.
a-f, Streptograptus sartorius ( Tornquist, 1881). A, BGS CAV185I. b-e, BGS SPT173. c, BGS SPT177. F, BGS
JZ2692. a, f, lower griestoniensis Biozone; stream section (SO 08436847) E of Abbeycwmhir. b-e, crispus
Biozone, sartorius Subzone; quarry (SN 997 57 477) 2 km S of Pant-y-dwr. g-k, Streptograptus aft. sartorius
(Tornquist, 1881 ); all crispus Biozone, sartorius Subzone; track cutting (SN 89007777) W of Llangurig. g, BGS
DJ8476. h, i, k, BGS DJ8502. j, BGS DJ8488. l-p, Torquigraptus pragensis pragensis (Pfibyl, 1943); all lower
griestoniensis Biozone; stream section (SO 08436847) E of Abbeycwmhir. l, BGS JZ2683; proximal to mesial
part of rhabdosome. m, part of l. n, same specimen as l; more distal part, o, counterpart of part of n. p, BGS
JZ2682; distal fragment, ventral aspect, q, Torquigraptus pragensis ruzickail (Pfibyl, 1943); BGS DJ7179,
upper crispus Biozone; Afon Ystwyth (SN 84227559).

a-k, x 20 (largest scale bar represents 2 mm); l, n, x 5 (smallest scale bar represents 2 mm); m, o-q, x 10

(intermediate scale bar represents 2 mm).

386

PALAEONTOLOGY, VOLUME 37

text-fig. 6. For legend see opposite.

ZALASIEWICZ : LATE LLANDOVERY GRAPTOLITES

387

apertures face diagonally, towards the outside of the spiral. This type of structure is not
torquigraptid (see below); it may be more closely related to that of Monograptus spiralis (see Lenz
and Melchin 1989) though the apertures are non-spinose.

SYSTEMATIC PALAEONTOLOGY

Order graptoloidea Lapworth, 1873
Suborder virgellina Fortey and Cooper, 1986
Family glyptograptidae Fortey and Cooper, 1986
Subfamily monograptinae Freeh, 1897
Genus streptograptus Yin, 1937 emend. Loydell, 1990

Type species. (Designated by Loydell 1990; see Loydell and Chen 1991); Graptolithus plumosus Baily, 1871 ;
from the Llandovery of Tievesvilly, County Down, Northern Ireland.

Streptograptus white i sp. nov.

Text-figure 3o-q

Derivation of name. After Dr D. E. White, who collected much of the material of this species.

cf. 1931 Monograptus nodifer, Haberfelner, p. 136, pi. 11, fig. la-c.

cf. 1982 Monograptus barrandei group Elies & Wood, Strachan, p. 163, fig. 2a-d.

Holotype. BGS DJ7612, from uppermost turriculatus Biozone ( carnicus Subzone) strata exposed beside
Claerwen reservoir (SN 8508 6415) (Text-fig. 3p-q).

Horizon and localities. Rare specimens, all from the carnicus Subzone of the turriculatus Biozone of
central Wales.

Diagnosis. A long, slowly expanding streptograptid with gentle, persistent ventral curvature.
Prothecae approximately parallel-sided in the proximal and mesial parts of the rhabdosome, and
distally narrowing in the distal part of the rhabdosome.

Description. A proximal end with sicula has not been found. Total rhabdosome length was probably
considerably greater than the maximum of 50 mm observed. All fragments show a persistent, gentle ventral
curvature. The most proximal fragment is 0 2 mm wide, with a 2TRD of 2 0 mm, and shows no perceptible
expansion over 15 mm. The holotype expands from 0 35 mm to 0 7 mm in 35 mm, the 2TRD decreasing from

text-fig. 6. Graptolites from the griestoniensis Biozone, a-i, Torquigraptus pergracilis ? (Boucek, 1931); all
lower griestoniensis Biozone; Nant Cerrigyrhdyg (SN 8501 7050), 10 km NW of Rhayader, a, BGS SPT2041 ;
b, detail of a. c, BGS SPT2053. d, detail of c. e, BGS SPT2088. f, BGS SPT2046. G, detail of f. h, BGS
SPT2050. i, BGS SPT2056; specimen with sicula? j-o, Streptograptus aff. loydelli Storch and Serpagli, 1993;
lower griestoniensis Biozone; landslip section (SN 924749) N of Rhayader, j, BGS JZ942. k, BGS JZ956. f,
BGS JZ945. m, BGS JZ964. n, BGS JZ971. o, BGS JZ930. p, Torquigraptus ex. gr. pragensisl (Pfibyl, 1943);
BGS JZ3989; upper griestoniensis Biozone; track section (SSN 85457831) 6 km WSW of Llangurig, Powys;
distal fragment, q-t, Monoclimacis cf. griestoniensis of Elies and Wood, 1911; lower griestoniensis Biozone;
stream section (SO 08436847) E of Abbeycwmhir. q, BGS JZ2653; distal fragment, r, BGS CAV1853;
proximal part, s, BGS JZ2652. t, BGS CAV1853; distal part of r. u, Monoclimacis griestoniensis griestoniensis
(Nicol, 1850); BGS DJ7517; upper griestoniensis Biozone; quarry (SN 92587994), E of Llangurig near-

proximal to mesial fragment.

a, c, e-f, h-u, x 10 (smaller scale bar represents 2 mm); d, g, x 20 (larger scale bar represents 2 mm).

388

PALAEONTOLOGY, VOLUME 37

text-fig. 7. For legend see opposite

ZALASIEWICZ : LATE LLANDOVERY GRAPTOLITES

389

2-1 mm to 1-6 mm over that distance. Other mesial fragments show slightly more widely spaced thecae, with
2TRDs of up to 2-4 mm. The thecae are of streptograplid type. Proximally and mesially, the prothecae are
approximately parallel-sided or show slight distal expansion, and comprise approximately half the total
rhabdosome width. Distally, prothecae show distal contraction, giving this part of the rhabdosome a
Tuncinate’ outline.

Discussion. Monograptus nodifer ( sensu Haberfelner, 1931, also from the upper turriculatus Biozone,
of the Carnic Alps) appears similar, differing only in having more closely set thecae, with a 2TRD
of c 1-5 mm. M. nodifer s.s ., however, is unrelated to the species being described, having radically
different thecal morphology (Rickards et cd. 1977). Streptograptusl anguinus anguinus (Pribyl.
1941), a spiralis Biozone species, and Streptograptusl anguinus linearis (Chen, 1984) from the
turriculatus Biozone of China have a generally similar rhabdosome form; both differ from S. white i
in having parallel-sided prothecae throughout and slightly lower profile metathecae which are
slightly more widely spaced both proximally and distally.

This rare species is useful stratigraphically in central Wales, having been found only in strata
assigned to the uppermost turriculatus Biozone (carnicus Subzone). It may also have been recorded,
as ‘M. barrandei group’, from uppermost turriculatus Biozone or lower crispus Biozone strata in
south-east Scotland, associated with Monoclimacisl galaensis (Strachan 1982).

Genus torquigraptus Loydell, 1 993/?

Type species. Original designation; Graptolithus Proteus Var. plana Barrandc, 1850; from the linnaei Biozone
of Zelkovice, Bohemia.

Discussion. This genus embraces monograptids which possess relatively simple thecal apertures
which are laterally twisted towards the reverse (right) side of the rhabdosome. This phenomenon
was first noted by Linnarsson (1881) for his species Monograptus dextrorsus from the turriculatus
Biozone of Sweden, and then by Tornquist (1899) for his species Monograptus denticulatus from the
Aeronian Diplograptus folium Biozone of Scania. The significance of these early findings was not
widely appreciated, however, and convincing further evidence of this morphological feature only
emerged relatively recently. In particular, Bjerreskov (1975) recognized thecal torsion in a number
of Llandovery monograptids from Bornholm: Monograptus planus , M. proteus , M. tullbergil, and
? Diver sograptus sp. and Melchin ( 1989) showed this feature in M. dextrorsus n. ssp. from Cornwallis
Island, Canada, while Loydell ( 1993/5) has demonstrated it in the late Aeronian to early Telychian
species M. involutus Lapworth, 1876, M. contortus Perner, 1897, and M. cavei Loydell, 19936 from
west Wales.

Torquigraptus is restricted to species with fairly simple apertures. Monograptus spiralis (see Lenz
and Melchin 1989), M. contortus (see Loydell 19936) and Spirograptus turriculatus (see Melchin and

text-fig. 7. Graptolites from the griestoniensis Biozone, a-i, Monograptus pseudocommunis sp. nov.; a-f, all
lower griestoniensis Biozone; from Nant Ccrngyrhydyg (SN 85017050), 10 km NW of Rhayader, a, BGS
SPT2034 ; latex cast of holotype. b, BGS SPT2088. c, detail of b. d, BGS SPT2050. e, BGS SPT2079. f, BGS
SPT2067. G— i, probably lower griestoniensis Biozone; specimens collected by R. O. Roberts (1929) from
'Pentre Brook’ (= Pandy Brook), g, ROR63 (R. O. Roberts collection, Cambridge Univ.). h-i, ROR570
(R. O. Roberts collection); ventrolateral and lateral aspects respectively, j, Torquigraptus tullbergi cf.
spiraloides (Pribyl, 1943); BGS JZ4496 a; lower griestoniensis Biozone, section (SO 05877280); N of
Abbeycwmhir. k m, Torquigraptus tullbergi spiraloides (Pribyl, 1945); respectively k-l, BGS GSM 1 1801 and
m, BGS Geol. Soc. Coll. 6951 ; griestoniensis Zone; Grieston Quarry, Innerleithen, Scotland. N, Monograptus
clintonensis (sensu Hall, 1852); BGS JZ2710; lower griestoniensis Biozone; stream section (SO 08436847) E of
Abbeycwmhir. o, Monograptus sp. 1 , BGS JZ2707 ; lower griestoniensis Biozone ; stream section (SO 08436847),

E of Abbeycwmhir.

a-b, d-g, j-n, x 10 (smaller scale bar represents 2 mm); c, h-i, o, x 20 (larger scale bar represents 2 mm).

PALAEONTOLOGY, VOLUME 37

text-fig. 8. For legend see opposite.

ZALASIEWICZ : LATE LLANDOVERY GR APTOLITES

391

Lenz 1986) possess laterally twisted thecae, but also more complex apertures with lateral spines or
flanges. Their relationship to Torquigraptus is uncertain. Cyrtograptus species, also possess laterally
twisted, and more complex, flange-bearing or spinose apertures (e.g. Lenz and Melchin 1989); this
genus, possibly polyphyletic (Lenz and Melchin 1989) may in part have descended from
Torquigraptus species. Torquigraptus species have so far been recognized from the Aeronian and
Telychian stages (see Loydell for further details). They are mostly, but not exclusively, spirally
coiled or dorsally curved.

Torquigraptus pergracilis ? (Boucek, 1931).

Text-figure 6a-i

71931 Monograptus pergracilis sp. nov., Boucek, p. 302, text-fig. 106.

71933 Diversograptus 7 pergracilis (Boucek); Boucek, p. 70, pi. 6, figs 5-6, text-fig. 176.

71952 Diversograptus pergracilis (Boucek); Munch, p. 132, pi. 44, fig. 2 a-c.

71953 Diversograptus capillaris pergracilis (Boucek); Boucek and Pfibyl, pp. 500, 561, pi. I, fig. 4, text-
fig. 2, figs 1-6.

Material. Between ten and fifteen fragmentary rhabdosomes from the lower part of the griestoniensis Biozone,
10 km NW of Rhayader, Wales; SN 8501 7050. One specimen from strata of probable lower griestoniensis
Biozone age collected by R. O. Roberts (1929) from 'Pentre Brook’ (probably Pandy Brook), near
Abbeycwmhir, Powys, Wales.

Description. Fragments at least 16 mm long; all possess a slight dorsal curve. One fragment (Text-fig. 6l) shows
a possible proximal end. The apparent sicula is at least 0-8 mm long. Width increases gradually from 0-2 mm
proximally to a maximum of 03 mm. Thecal shape is consistent throughout. Prothecae are long and nearly
parallel-sided; metathecae are small, inconspicuous and appear to be consistently twisted to the right. 2TRDs
increase slightly from 2-35 mm proximally to 2-7-3-2 mm distally.

Remarks. The almost straight and very slender rhabdosome suggests comparison with pergracilis.
P. Storch has kindly re-examined Boucek's original material and furnished the following description
of the poorly preserved type material from Bohemia: ’Rhabdosome is very slender, either straight
or weakly curved (both ventral and dorsal curvature observed). Elongated tube-like thecae overlap
for about one-tenth to one-eighth of their length. Metathecae are hooked and, possibly (laterally)
twisted. Probably they face proximolaterally Thecal height is 0-3 mm (0-25-0-35 mm), and 2TRD
is 2-4— 2-8 mm (mostly 2-6 mm)’. Thus, the Welsh specimens may resemble the type Bohemian
material in thecal morphology as well as in overall rhabdosome shape.

Boucek’s species, however, has been recorded from higher horizons ( crenulata and spiralis
Biozones of Bohemia; Boucek 1953); some of the material showed thecal cladia (Boucek and Pfibyl
1953), a feature not seen in the Welsh specimens. If the Welsh material is pergracilis , then the placing

text-fig. 8. Graptolites from the crenulata and spiralis Biozones, a-b, Monograptus all. falx (Suess, 1851);
probably spiralis Biozone; quarry (SO 00927707) NW of Bwich-y-sarnau. a. BGS DJ9943; b, BGS DJ9968.
c-d, Pristiograptusl aff. initialis (Kirste, 1919); spiralis Biozone; quarry (SO 00927707) NW of Bwlch-y-
Sarnau. c. BGS SPT9. d, BGS DJ9916. e-f, Torquigraptus tullbergi tullbergi (Boucek. 1931); E, BGS JZ4232;
probably crenulata Biozone; track section (SO 08177417) E of Bwlch-y-Sarnau. f, counterpart of e; BGS
JZ4233. G, Pristiograptus initialis (Kirste, 1919); BGS JZ4256; probably crenulata Biozone; quarry (SO
08177417) E of Bwlch-y-Sarnau. h-k, Monoclimacis cf. crenulata (sensu Elies and Wood, 1911); probably
crenulata Biozone; quarry (SO 08177417) E of Bwlch-y-Sarnau. H, BGS JZ4218. i, BGS JZ4191. j, BGS
JZ4178. K, BGS JZ4178. L-M, Monoclimacis griestoniensis nicolil Rickards, 1965; probably spiralis Biozone;
quarry (SO 00927707) NW of Bwlch-y-Sarnau. L, BGS DJ9970; m, (counterpart of l) BGS DJ9971). n-q,
Monoclimacis linnarssoni (Tullberg, 1883); horizon and locality as for l-m. n, BGS SPT20. o, BGS DJ9959.
p, BGS DJ9997. q, BGS SPT20. r, Monograptus priodon (Bronn); BGS JZ4218; probably crenulata Biozone;

quarry (SO 08177417) E of Bwlch-y-Sarnau.

All x 10 (larger scale bar represents 20 mm), except Q, x 5 (smaller scale bar represents 2 mm).

392

PALAEONTOLOGY, VOLUME 37

of this taxon by Boucek and Pribyl (1953) as a subspecies of capillaris is incorrect, as the latter
species has symmetrical thecae (Zalasiewicz, unpublished).

A seemingly allied, but more robust taxon with more widely spaced thecae (2TRD c. 3-8 mm)
from crispus Biozone strata in central Wales (Text-figs 4m-o) shows lateral twisting of the
metathecae. This may be related to German material assigned to pergracilis in Boucek’s collection,
which may be from the crispus Biozone (M finch 1952), which has 2TRDs of 34-4-2 mm (P. Storch,
pers. comm.).

Genus Monograptus Geinitz, 1842

Type species. By subsequent designation (Bassler 1915, p. 822) Lomatoceras priodon Bronn, 1835, p. 56, pi. 1,
fig. 13; from the Silurian of Germany.

Monograptus crispus Lapworth, 1876

Text-figure 4a-h

1876 Monograptus crispus sp. nov., Lapworth, p. 503, pi. 120, fig. la-c.

1913 Monograptus crispus Lapworth; Elies and Wood, p. 456, pi. 45, fig. 6 a-f; text-fig. 314 a-c.

1951 Monograptus ( Globosograptus ) crispus Lapworth; Boucek and Pribyl, p. 192, pi. 1, fig. 1-7; pi.

2, figs 1-3.

1975 Monograptus crispus Lapworth; Hutt, p. 84, pi. 11, figs 8-9, text-fig. 25, fig. 5.

1985 Prochnygraptus crispus (Lapworth); Pribyl and Storch, p. 62, pi. 1, figs 1-2; pi. 2, ?1, ?4.

Lectotype. BLT 1648, figured Lapworth 1876, pi. 20, fig. 7; from the Gala Beds of Meigle Quarry, Scotland.

Description. The material consists of a proximal, dorsally curved portion 0-2-0-25 mm across, with a 2TRD of
c. 2-0 mm; a middle portion which is approximately straight and which varies considerably in length, with a
width of 0-25-0-35 mm and 2TRDs of 3-CML0 mm; and a ventrally curved distal portion up to 0-9 mm wide,
with 2TRDs of 2 0-2-2 mm. The thecae show a progressive distal increase in the amount of coiling. Thecal
overlap is negligible throughout and there are no prothecal folds. Proximally, the prothecae are long, and
gently distally expanding. The metathecae are coiled through c. 270° (measuring from the local stipe axis). The
aperture is gently flared and faces, but is not tightly pressed against, the ventral prothecal wall (Text-fig. 4c).
Mesially, the prothecae remain narrow, and become longer. The metathecae are more tightly coiled than in the
proximal portion, through c. 360°. The apertures are flared, possessing a distinct lateral flange; they are tightly
pressed against the initial part of the metatheca (Text-fig. 4d-f). Distally, the prothecae shorten and become
more rapidly distally expanding, the coiled metatheca forming an isolated lobe. The amount of coiling here is
c. 450°, the flared apertures facing ventrally and being tightly adpressed against the dorsal-facing wall of the
preceding part of the metatheca (Text-fig. 4a-b, g-h).

Discussion. The only previous detailed description of the thecal structure of M. crispus is in an
unpublished Ph.D. thesis (Howe 1982). The main difference of these observations from those of
Howe lies in the recognition of the progressive increase in coiling from the mesial to the distal part
of the rhabdosome. The affinities of this distinctive thecal type is unknown, though an ancestral
relationship to the folded metathecae of M. knockensis and M. singularis , inferred by Pribyl and
Storch (1985), is possible. The coiled metathecae with their flared apertures suggest an affinity with
the streptograptids, but the lack of prothecal folds and the progressive increase in the amount of
coiling may indicate otherwise.

Monograptus aff. crispus
Text-figure 3j-l

Material. Very rare specimens in strata of latest turriculatus Biozone ( carnicus Subzone) to early crispus
Biozone in age.

Discussion. This rare taxon appears to predate slightly M. crispus (having been found together with
Streptograptus storchi , which has not yet been found to co-exist with M. crispus s.s. in Wales) and

ZALASI EWICZ : LATE LLANDOVERY GR APTOLITES

393

to overlap with the lower part of its range. It is characterized by more closely-set thecae than is
typical of M. crispus , with 2TRDs in the range of 1- 1-1-5 mm. It is also somewhat less robust, the
three distal fragments found not exceeding 0-5 mm in width.

Monograptus pseudocommunis sp. nov.

Text-figure 7a- i

Derivation of name. From the superficial resemblance of this taxon to the Aeronian species Monograptus
communis.

Holotype. SPT 2034 (Text-fig. 7a); from the early griestoniensis Biozone in a stream section (SN 8501 7050),
10 km NW of Rhayader, Wales.

Paratypes. Five additional specimens from the type locality. In addition, two well-preserved fragments from
strata of probable lower griestoniensis Biozone age from ‘Pentre Brook’ (probably Pandy Brook), near
Abbeycwmhir, Powys were collected by R. O. Roberts (1929).

Diagnosis. Rhabdosome strongly dorsally curved, maximum width TO mm; proximal 2-3 thecae
slender, elongated; distal thecae triangulate, hooked, symmetrical, with slightly laterally expanded,
ventrally facing apertures.

Description. The short rhabdosome has a dorsal curve, accentuated mesially. The proximal end is gracile,
c. 0-1 5 mm wide with a proximal 2TRD of 2 0 mm. It widens rapidly from theca 2 (0-25 mm) to theca 4 (0-6 mm),
and thereafter more gradually to 10 mm. Distal 2TRDs are T5-T9 mm. Distal thecae are triangulate, hooked,
and symmetrical. Two well-preserved thecae (Text-fig. 7h-i) show that the apertural portions of the metathecae
are slightly expanded, and that the dorsal apertural wall extends further than the ventral apertural wall, so that
the aperture faces ventrally.

Discussion. This species closely resembles the Aeronian species M. communis in rhabdosonial shape
and thecal structure, differing only in being narrower distally (maximum width TO mm as opposed
to T3-T4 mm) and having ventrally facing apertures. This may be a true triangulate monograptid;
but, as the triangulates, otherwise, appear to have disappeared sometime around the beginning of
the Telychian, it is possible that derivation was from Torquigraptus. If so, it would constitute an
example of iterative evolution - independently re-creating the communis morphology - and also
reverse evolution, ‘straightening out’ the laterally twisted thecae.

Acknowledgements . The author is grateful to colleagues in the BGS Wales Research Group, particularly Dr
R. A. Waters, and Dr A. W. A. Rushton for much help and encouragement, to Dr D. K. Loydell of the
University of Wales, Aberystwyth, for useful discussions and comments on the manuscript, and to Dr P. Storch
of the Czech Geological Survey for the furnishing of useful comparative data. This paper is published with the
permission of the Director of the British Geological Survey (NERC).

REFERENCES

baily, w. h. 1871. Palaeontological remarks. 22-23. In traill, w. a. and egan, f. w. Explanatory memoir to
accompany sheets 49. 50 and part of 61 of the maps of the Geological Survey of Ireland including the country
around Downpatrick, and the shores of Dundrum Bay and Strangford Lough, County of Down. Alexander
Thom, Dublin and London, 71 pp.

barrande, j. 1850. Graptolites de Boheme. vi. Prague, 71 pp.

bassler, r. 1915. Bibliographic index of American Ordovician and Silurian fossils. Bulletin , United States
National Museum , 92, 1-718.

bjerreskov, m. 1975. Llandoverian and Wenlockian graptolites from Bornholm. Fossils and Strata. 8, 1 94.
pis 1-13.

394

PALAEONTOLOGY, VOLUME 37

bjerreskov, m. 1981. Silurian graptolites from Washington Land, western New Greenland. Bulletin Gronkmds
Geologiske Undersogelse, 142, 1—71, pis 1-6.

boucek, b. 1931. Predbezna zprava o nekterych novych druzich graptolitu z ceskeho gotlandienu.
Communication preliminaire sur quelques nouvelles especes de graptolites provenant du Gothlandian de la
Boheme. Vestnik Stdtniho Geologickeho Ustavu Republiky Ceskoslovenske , 7, 293-313.

1933. Monographic der Obersilurischen Graptolithen aus der Familie Cyrtograptidae. Prdce Geologicko-
Palaeontologiceho Ustavu Karlovy University v Praze, 1, 1-84, jrls 1-7.

1953. Biostratigraphy, development and correlation of the Zelkovice and Motol Beds of the Silurian of
Bohemia. Sbornik ustfedniho ustavu Geologickeho , 20, 473 484.

and bribyl, a. 1942. Uber bohmische Monograpten aus der Untergattung Streptograptus Yin.
Mitteilungen der Tschechischen Akademie der Wissenschaften , 52, 1-23, pis 1-3.

1951. On some slender species of the genus Monograptus (Geinitz), especially of the subgenera
Mediograptus and Globosograptus. Bulletin International de L' Academic Tcheque des Sciences, 52, 185-216,
pis 1-3.

1953. O rodu Diversograptus Manck z ceskeho siluru. [On the genus Diversograptus Manck from the
Silurian of Bohemia]. Sbornik Ustfedniho Ustavu Geologickeho, 20, 485-576, pis 1-3.
carruthers, w. 1867. On graptolites. Appendix D. 538-541 in Murchison, r. i., Siluria, 4th edition, John
Murray, London, 566 pp.

cave, r. 1979. Sedimentary environments of the basinal Llandovery of mid-Wales. 517-526. In Harris, a. l.,
Holland, c. h. and leare, b. E. (eds). The Calcdonides of the British Isles - reviewed. Geological Society,
Special Publication, 8, 768 pp.

chen, xu. 1984. Silurian graptolites from southern Shaanxi and northern Sichuan with special reference to the
classification of the Monograptidae. Palaeontologica Sinica, New Series B, 166, No. 20, 1-102, pis 1-19.
DAVIES, J. R., FLETCHER, C. J. N., WATERS, R. A., WILSON, D., WOODHALL, D. and ZALASIEWICZ, J. A. in press. The
geology of the Rhayader and Llanilar districts. Memoir of the British Geological Survey (in press).

Davies, K. a. 1929. Notes on the graptolite faunas of the Upper Ordovician and Lower Silurian. Geological
Magazine, 66. 1-27.

elles, G. L. and wood. E. M. R. 1901-1918. A monograph of British graptolites. Monograph of the
Palaeon tographical Society. Part 8, 191 I, 64 (316), 359^114, pis 36-41 ; Part 9, 1913, 66 (323) 415-486, pis
42-49.

fortey, r. a. and cooper, r. a. 1986. A phylogenetic classification of the graptoloids. Palaeontology, 29,
631-654.

frech. F. 1897. Lethaea Geognostica. Tlieil 1 . Lethaea Paleozoica, 1.11. Graptolithen, 544-684. Schweizerbach,
Stuttgart.

geinitz, h. b. 1842. Uber die Graptolithen. Neues Jahrbuch fur Mineralogie, Geografie, Geologie und
Petrefakten- Kunde, 1842, 697-701. pi. 10.

gortanl M. 1923. Contribuziom alio studio del Paleozoico Carnico. Parte VII, Graptoliti del Monte
Hochwipel. Palaeontographica Italica , 29, 1-24, pi. 1.

iiaberfelner, e. 1931. Graptolithen aus den Obersilur der Karnischen Alpen. I. Teil: Hochwipel, Nordseite.

Sitzungsberichte. Akademie der Wissenschaften in Wien (1), 140, 89-168, pis 1-3.
hall, J. 1843. Geology of New York. Report on Fourth Geological District.

1852. Palaeontology of New York. Volume 2, containing descriptions of the organic remains of the Lower
Middle Division of the New- York System. C. van Benthuysen, Albany, New York, 362 pp.
howe, M. p. a. 1982. Upper Llandovery graptolites and stratigraphy of the northern Oslo region. Unpublished
Ph.D. thesis. University of Cambridge.

1983. Measurement of thecal spacing in graptolites. Geological Magazine, 120, 635-638.
hu ll j. e. 1975. The Llandovery graptolites of the English Lake District. Part 2. Palaeontographical Society
Monograph, 129 (542), 57-137, pis 1 1 -26.

rickards. r. b. and skevington, d. 1970. Isolated Silurian graptolites from the Bollerup and Klubudden
stages at Dalarna, Sweden. Geologica et Palaeontologica, 4, 1-23, pis 1-3.
jones, o. t. 1909. The Hartfell-Valentian succession in the district around Plynlimon and Pont Erwyd
(Cardiganshire). Quarterly Journal of the Geological Society, London. 65, 463-537, pis 24-25.
kirste, E. 1919. Die Graptolithen des Altenburger Ostkreises. Mitteilungen Osterlande, 16, 60-222, pis 1-3.
lapworth, c. 1873. Notes on the British Graptolites and their allies. 1. On an improved classification of the
Rhabdophora. Geological Magazine, (I), 10, 500-504, 555-560, Table 1.
lapworth, c. 1876. On Scottish Monograptidae. Geological Magazine, (2), 3, 308-21, 350-60, 499-507,
544-52, pis 10-13, 20.

ZALASIEWICZ : LATE LLANDOVERY GRAPTOLITES

395

lapworth, H. 1900. The Silurian sequence of Rhayader. Quarterly Journal of the Geological Society , London ,
56, 67-137, pis 6-7.

lenz, A. c. and melchin, M. J. 1989. Monograptus spiralis and its phylogenetic relationship to early
cyrtograplids. Journal of Paleontology, 63. 341-8.

linnarsson, G. 1881. Graptolitskiffrar med Monograptus turriculatus vid Klubudden nara Motala. Geologiska
Foreningens i Stockholm Forhandlingar , 5, 503-527, pis 22-23.
loydell, d. k. 1990a. On the type material of the Llandovery (Silurian) graptoloid Monograptus pseudobecki
Boucek and Pfibyl, 1942 and the species' biostratigraphical importance. Scottish Journal of Geology , 26,
119-123.

— 19906. On the graptolites described by Baily (1871) from the Silurian of Northern Ireland and the
genus Streptograptus Yin. Palaeontology , 33, 937-943.

1991. The biostratigraphy and formational relationships of the upper Aeronian and lower Telychian
(Llandovery, Silurian) formations of western mid-Wales. Geological Journal , 26, 209-244.

1992m Middle and Upper Llandovery graptolites from western mid-Wales. Part 1. Monograph of the
Palaeontograpliical Society, 146 (590), 1-55, pi. I.

— 19926. Graptolithus clintonensis (currently Monograptus clintonensis ; Graptolithina) : proposed attribution
to Hall, 1852, and designation of a lectotype. Bulletin of Zoological Nomenclature , 49, 43-45.

— 1993a. Worldwide correlation of Telychian (Upper Llandovery) strata using graptolites. 323-340. In
hailwood, e. A. and kidd, r. E. (eds). High resolution stratigraphy in modern and ancient marine sequences.
Geological Society Special Publication No. 70, 360 pp.

19936. Middle and Upper Llandovery graptolites from western mid-Wales. Part 2. Palaeontographical
Society Monograph, 147 (592), 56-180, pis 2-5.

BULL, E. E. and storch, p. 1992. Monograptus crenulatus (currently Monoclimacis crenulata ;
Graptolithina): proposed attribution of the specific name to Elies and Wood, 1911, and proposed
designation of a lectotype. Bulletin of Zoological Nomenclature , 49, 46-48.

— and cave, r. 1993. The Telychian (Upper Llandovery) stratigraphy of Buttington Brick Pit, Wales.
Newsletters in Stratigraphy, 29, 91-103.

— and chen, x. 1991. Streptograptus Yin, 1937 (Graptolithina): proposed designation of Graptolithus
plumosus Baily, 1871 as the type species. Bulletin of Zoological Nomenclature, 48, 236-237.

— and zhao, y. h. 1990. Studies on some monograptids using the scanning electron microscope. Acta
Palaeontologica Sinica, 29, 331-336, pis 1-3.

melchin, m. j. 1989. Llandovery graptolite biostratigraphy and paleoecology. Cape Phillips Formation,
Canadian Arctic Islands. Canadian Journal of Earth Sciences, 26, 1726-1746.

— and lenz, a. c. 1986. Uncompressed specimens of Monograptus turriculatus (Barrande, 1850) from
Cornwallis Island, Arctic Canada. Canadian Journal of Earth Sciences, 23, 579-582.

munch, a. 1952. Die Graptolithen aus dem anstehenden Gotlandium Deutschlands und der Tschechoslovakei.
Geologica, 7, 1-157, pis 1-62.

nicholson, h. a. 1868. On the graptolites of the Coniston Flags; with notes on the British species of the genus
Graptolites. Quarterly Journal of the Geological Society, London , 24, 521-545, pis 19-20.

— 1872. A monograph of the British Graptolidae. Part 1. General Introduction. Blackwood and Sons,
Edinburgh and London, xiii+ 133 pp.

NICOL, J. 1850. Observations on the Silurian Strata of the south east of Scotland. Quarterly Journal of the
Geological Society , London , 6, 53-65.

perner, j. 1897. Etudes sur les graptolites de Boheme , Illieme Partie. Monographic des Graptolites de I'Etage E.
Section a. 1-25, pis. 9-13, Prague.

pribyl, a. 1940. Revise ceskych graptolitu rodu Monoclimacis Freeh. Rozpravy II, Truly Ceske Akademie, 50
(23). 1-19, pis 1-3.

— 1941 . O ceskych a cizych zastupcych rodu Rastrites Barrande 1850. Rozpravy II, Tfidy Ceske Akademie,

51 (6), 1-22, pis 1-3.

1943. Prfpisek k poznani nemeckych zastupcu rodu Rastrites Barr. Rozpravy II, Tfidy Ceske Akademie,

52 (4), 1-10, pi. 1.

1945. The Middle European Monograptids of the genus Spirograptus. Giirich. Bulletin International de
f Academie Tcheque des Sciences, 45, 185-231, pis 1-1 1.

— and storch, p. 1985. Prochnygraptus gen. n. (Graptolithina, Monograptidae) from the Middle and Upper
Llandovery of Europe. Vestnik Ustfedniho Ustavu Geo/ogickeho, 60, 159-164, pis 1-2.

richter, R. 1871. Aus dem thuringischen Schiefegebirge. Zeitschrift der Deutschen Geologischen Gesellschaft,
23, 231-256, pi. 5.

396

PALAEONTOLOGY, VOLUME 37

rickards, R. b. 1965. New Silurian graptolites from the Howgill Fells (Northern England). Palaeontology. 8.
247-271.

1970. Llandovery graptolites from the Howgill Fells. Paleontographical Society Monograph , 123 (524),
1-108. pis 1-6.

- 1976. The sequence of Silurian graptolite faunas in the British Isles. Geological Journal. 1 1, 153-188.

- 1989. Exploitation of graptolite cladogenesis in Silurian biostratigraphy. 183-200. In Holland, c. h. and
bassett, m. G. (eds). A global standard for the Silurian System. National Museum of Wales, Geological Series
No. 9., Cardiff, 325 pp.

- hutt, j. E. and berry, w. B. N. 1977. Evolution of the Silurian and Devonian graptoloids. Bulletin of the
British Museum (Natural History ), Geology Series. 28, 1-120, pis 1-6.

Roberts, R. o. 1929. The geology of the district around Abbey-cwmhir (Radnorshire). Quarterly Journal of the
Geological Society, London, 85, 651-676.

schauer, m. 1971. Biostratigrafie und Taxonomie der Graptolithen des tieferen Silurs unter besonderer
Berucksichtung der tektonischen Deformation. Freiburger Forschungshefte , C373, Palaontologie, 1-185.
storch, p. and serpagli, e. 1993. Lower Silurian graptolites from southwestern Sardinia. Bollettino della
Societd Paleontologica Italiana, 32, 3-57.

strachan, i. 1982. The distribution of graptolites in south-east Scotland, with notes on the fauna of Old
Cambus Quarry. Scottish Journal of Geology, 18, 157 -165.
sudbury, m. 1958. Triangulate monograptids from the Monograptus gregarius Zone (Lower Llandovery) of the
Rheidol Gorge (Cardiganshire). Philosophical Transactions of the Royal Society of London, Series B241,
485-554.

suess, e. 1851. Uber bohmischc Graptolithen. Naturwissenschaftliche Abhandlungen, 4. 87-134, pis 7-9.
tornquist, s. L. 1881. Om nagra nya graptoliter fran Dalame. Geologiska Foreningens i Stockholm
Forhandlingar, 5, 434-445.

1883. Oerversigt ofver bergbygnaden inom Siljansomradet i Dalarne. Sveriges Geologiska Undersoknung ,
Series C, 57, 1-59.

1892. Undersokningar ofver Siljansomradets graptoliter. II. Lunds Universitets Arsskrifter, 26, 1-33,
pis 1-2.

1899. Researches into the Monograptidae of the Scanian Rastrites Beds. Lunds Universitets Arsskrifter,
35. 1-25, pis 1-4.

tullberg, s. a. 1883. Skanes graptoliter. II. Graptolitfaunorna i Cardiolaskiffern och Cyrtograptuskiffrarne.

Sveriges Geologiska Undersoknung, Series C, 55, I 43.
wood, e. m. r. 1906. The Tarannon Series of Tarannon. Quarterly Journal of the Geological Society, London,
62, 644-701, pis 47-48.

Wilson, D. w. R. 1954. The stratigraphy and palaeontology of the Valentian rocks of Cautley. (Yorks W.R. ).
Unpublished Ph D. thesis. University of Birmingham.

yin, T. H. 1937. Brief description of the Ordovician and Silurian fossils from Shihtien. Bulletin of the Geological
Society of China, 16, 281-302, pis 1-2.

zalasiewicz, j. a. 1990. Silurian graptolite biostratigraphy in the Welsh Basin. Journal of the Geological
Society, London, 147, 619-622.

JAN ZALASIEWICZ
British Geological Survey

Typescript received 2 June 1993 Key worth

Revised typescript received 18 November 1993 Nottingham NG12 5GG, UK

THE POSTC R AN I A L SKELETON OF THE EARLIEST
DICYNODONT SYNAPSID EODICYNODON FROM
THE UPPER PERMIAN OF SOUTH AFRICA

by B. S. RUBIDGE, G. M. KING Gild P. J. HANCOX

Abstract. Comparison of postcranial skeletal elements of the Upper Permian Eodicynodon oosthuizeni (the
earliest member of the Dicynodontia) and other dicynodonts, shows distinctive features of the Eodicynodon
humerus, scapula, femur and ilium which are diagnostic at the generic level, and which may therefore aid
stratigraphical studies.

Both the fore- and hind limbs and girdles show a less derived condition than in other dicynodonts. The forelimb
adopted a sprawling position and was supported by extensive postural musculature as evidenced by the broad
blade of the scapula where the serratus anterior and levator scapulae attached, and by the large ventral girdle
for attachment of the pectoralis and coraco-brachialis. The more distal insertions of the deltoideus and
pectoralis muscles on the delto-pectoral crest suggest that the characteristic slow and powerful forelimb action
of later dicynodonts was not yet in evidence in Eodicynodon. It appears that the hind limb adopted a sprawling
or semi-sprawling position. The femur lacks a developed trochanter major, and the ilium is without marked
anterior and posterior processes. The ilio-femoralis was predominantly an elevator, retraction of the femur
having been carried out by the caudi-femoralis and ventral musculature attached to the extensive pubo-
ischiadic plate.

Eodicynodon Barry, 1974, is the earliest and most primitive member of the Dicynodontia {sensu
Rubidge and Hopson 1990; = 'higher dicynodonts’ of King 1988), which were herbivorous
mammal-like reptiles abundant during the Permian and Triassic. Eodicynodon is a medium-sized
dicynodont with a total body length of about 450 mm, and it stood about 150 mm high at the
shoulder. The genus contains two species, E. oosthuizeni (Barry, 1964) and E. oelofseni (Rubidge,
1990c), and is known from the Late Permian Eodicynodon-Tapinocaninus Assemblage Zone
(Rubidge 1990c/), the lowermost biozone of the Abrahamskraal Formation of the Beaufort Group
in South Africa. Although the cranial morphology has been described in considerable detail (Barry
1974; Cluver and King 1983; Rubidge 1984; 19906) and analysed in functional terms (King et at.
1989) the postcranial skeleton has so far received no treatment. Well-preserved and -prepared
material of Eodicynodon oosthuizeni is now available, so this gap can be filled.

A detailed knowledge of the skeleton of Eodicynodon is important for two reasons. First, since
Eodicynodon is the least derived dicynodont known, an understanding of its postcranial skeleton is
necessary before a review of dicynodont postcranial function can be undertaken. Secondly, it is
important as part of a complete revision of the postcranial skeleton of the principal anomodont
genera, the aim of which is to recognize generic characteristics in postcranial elements. These
elements can then be used for biostratigraphical purposes.

In this paper the distinguishing features of the postcranial skeleton of Eodicynodon oosthuizeni are
described and compared with other described dicynodont postcrania ( Rohertia hroomiana (King
19816), Dicynodon trigonocephalus (King 1981//), Cistecephalus (Cluver 1978), Kingoria nowacki
(Cox 1959; King 1985)), and an undescribed specimen of Diictodon galeops (SAM K1633). Apart
from Rohertia , these forms occur in later assemblage zones than does Eodicynodon. Rohertia
overlaps with Eodicynodon in the Eodicynodon-Tapinocaninus assemblage zone, but is not found as
low in that zone as Eodicynodon (see below).

[Palaeontology, Vol. 37, Part 2, 1994, pp. 397— 408. |

© The Palaeontological Association

398

PALAEONTOLOGY, VOLUME 37

The study is based on several well-preserved specimens because no one specimen is complete.
Most specimens consist of the major parts of articulated skeletons found in association with cranial
material and we can therefore be confident of the generic identification of the individual skeletal
elements.

There is a small size range in the specimens. The femur of the largest specimen (NMQR 3153) is
c. 20 per cent longer than that of the smallest (NMQR 3155). Although there are indications of some
allometric changes (e.g. shaft diameter of the femur becomes proportionally greater with increasing
femur length) these do not appear to be very great.

MATERIAL STUDIED

Specimens of Eodicynodon used in this study from the South African Museum, Cape Town (SAM)
and the National Museum Bloemfontein (NMQR) are:

NMQR 2991 ; coracoid plate, humeri, femora, tibia, fibula, atlas and other vertebrae, ribs (Text-fig.
1a); from Modderdrift, Prince Albert, Cape Province, South Africa.

NMQR 2992; skull, lower jaw, scapulocoracoid, clavicles, interclavicle, radius, atlas-axis complex,
vertebrae and ribs; from Tuinkraal, Prince Albert, Cape Province, South Africa.

NMQR 3153; scapula, part of coracoid, humerus, radius, part of manus, femur, tibiae, fibulae,
vertebrae and ribs (Text-figs 2a-b); from Modderdrift, Prince Albert, Cape Province, South Africa.

NMQR 3154; occiput, left ramus of lower jaw, scapulae, clavicles, humeri, radii, ulnae, manus,
vertebrae and ribs (Text-figs 1b, 2c, 3, 4b); from Botterkraal, Prince Albert, Cape Province, South
Africa.

NMQR 3155; ilia, pubo-ischiadic plates, femur, tibiae, fibulae, pes, vertebrae and ribs (Text-fig.
5a-b); from Combrinckskraal, Prince Albert, Cape Province, South Africa.

NMQR 3156; most of the left manus; from Botterkraal, Prince Albert, Cape Province, South
Africa.

NMQR 3158; Posterior half of skull, scapula, humerus, radius, ulna, manus (Text-rig. 4a); from
Swartgrond. Rietbron, Cape Province, South Africa.

POSTCRANI AL SKELETON

Scapula

The scapula (Text-fig. 1a) has a well-marked acromion process clearly demarcated from the rest of the bone.
The blade is broad and not as constricted above the acromion process as in other dicynodonts. Dorsally the
blade fans out to form a dorsal edge which is considerably broader than in Dicynodon trigonocephalus ,
Kingoria , Robertia and Cistecephalus (Cluver 1978).

Coracoid and procoracoid

These two bones together form a large plate (Text-fig. I a). The procoracoid is roughly square and contains a
procoracoid foramen (Text-fig. I a; pc.f). Smaller foramina near the border with the scapula are probably due
to damage to the bone. The coracoid is a smaller bone which tapers posteriorly (Text-fig. 1a; cor). The glenoid
facet of the coracoid is well-developed, rounded and faces postero-medially. The coracoid plate is much larger,
compared with the size of the scapula, than that of Robertia and D. trigonocephalus , but is comparable with
that of Diictodon (SAM K1633).

RUBIDGE ET AL.\ EODICYNODON POSTCRANIUM

399

text-fig. 1. Scapula and clavicle of Eodicynodon. a, scapula of NMQR 2991. b, clavicle of NMQR 3154 with
proximal end in dorsal view. Part of distal end reconstructed from NMQR 2991. Scale represents 10 mm.

text-fig. 2. Humerus of Eodicynodon. A, NMQR 3153 in anterior view. B, NMQR 3153 in dorsal view,
c, distal end of NMQR 3154 in ventral view. Scale represents 10 mm.

Clavicle

The clavicle (Text-fig. 1b) is of typical dicynodont shape, being a slender rod of bone with both proximal and
distal ends expanded and oriented at right angles to one another. The expanded ends are markedly bigger than
in Dicynodon trigonocephalus and are comparable with those of Robertia.

400

PALAEONTOLOGY, VOLUME 37

text-fig. 3. Radius and ulna of Eodicynodon. NMQR 3154. a, left ulna lateral view, b, left ulna in medial
view, c, left radius in lateral view, d, left radius in medial view. Scale represents 10 mm.

Interclavicle

The interclavicle is not exposed fully in the specimens available but is overlain by the clavicles. It has a long
posterior projection and would presumably be T-shaped if fully exposed. This would approximate to the
condition in Robertia and contrast with the smaller, squarish interclavicle of D. trigonocephalus. No sternum
is present in the specimens available for study.

Humerus

The humerus (Text-fig. 2) is a robust bone. The proximal and distal ends are set at an angle of c. 80-90° to
one another, which is considerably greater than in the other dicynodonts used here for comparison. There is
not trace of an ectepicondylar foramen. This is in contrast to the condition in the slightly later form, Robertia ,
where a notch can be seen in the position of the ectepicondylar foramen (King 19816, fig. 4c). Both the
proximal and distal ends are wide compared with the shaft of the bone. The distal end has prominent rounded
condyles for articulation with the ulna and radius (Text-fig. 2c). The posterior edge of the entepicondyle bears
a rounded well-marked attachment area for the lower limb flexor muscles (Text-fig. 2b-c; fl.). The delto-
pectoral crest is large and distinctive (Text-fig. 2a; d.p.cr.). It takes the form of a triangular plate extending
from the proximal surface of the bone to c. 40 per cent the length of the bone. The proximo-ventral edge is
very thin. The distal end of the crest grades smoothly into the body of the bone. The form of the crest contrasts
with that of later forms such as Dicynodon. Here the crest is an oval plate of bone which extends much farther
down the shaft, occupying c. 60 per cent of the length of the bone (King 1981a, fig. 19c). The head of the
humerus (Text-fig. 2a-b; he.) is indistinct, in contrast to the ulnar and radial condyles (Text-fig. 2c; ul.c., ra.c.)
which are well-developed.

Radius and ulna

The radius and ulna (Text-fig. 3) are both slender bones compared with those of Diictodon and Dicynodon
trigonocephalus , and more similar to those of Robertia. The ulna is larger than the radius and has a well-marked
olecranon process. The bone is medio-laterally compressed. There are well-marked hollows on its anterior and

RUBIDGE ET AL.\ EODICYNODON POSTCRANIUM

401

text-fig. 4. Manus of Eodicynodon. a, right manus of NMQR 3158 in dorsal view, b, right manus of NMQR

3154 in ventral view. Scale represents 10 mm.

posterior surfaces, presumably for muscle attachment. The ends of the radius are noticeably expanded. Rugose
ridges along the length of the bone mark the division between extensor and flexor muscle groups.

Manus

The manus (Text-fig. 4) is broad with long slender ungual phalanges. Other phalanges are slightly longer than
broad. Metacarpals are not markedly longer than phalanges. The most complete manus (NMQR 3156), which
contains three virtually complete digits, indicates that the phalangeal formula was probably 2, 3, 3, 3, 3, as in
all Anomodontia (includes the families Galechiridae and Dicynodontidae sensu Rubidge and Hopson 1990)
including Patranomodon (Rubidge and Hopson 1990), which is the least derived anomodont known. The
terminal phalanges are long and claw-like. It is not possible to ascertain the number of carpals. As in other
dicynodonts, the manus is short and wide. This contrasts with the more slender foot found in Patranomodon
(Rubidge and Hopson, work in progress), and Galechirus (Brinkman 1981) where the metacarpals are
approximately twice as long as the phalanges. In Eodicynodon the lengths of the metacarpals and phalanges
are subequal.

Ilium

The ilium (Text-fig. 5) is small compared with the pubo-ischiadic plate, unlike the condition in Robertia ,
Dicynodon trigonocephalus or Kingoria where the two elements are subequal. The blade of the ilium has small
anterior and posterior processes. The anterior process (Text-fig. 5a; ant.p.) is not sharply demarcated from the
rest of the bone, but grades smoothly into the anterior edge as in Robertia. The ilia of Robertia and
Eodicynodon are very similar except that the ilium of Eodicynodon is taller owing to a longer neck between the

402

PALAEONTOLOGY, VOLUME 37

pu.isc.f .

text-fig. 5. Pelvic girdle of Eodicynodon. a, right side of pelvis of NMQR 3155 in lateral view, b, right
ischium of NMQR 3155 in ventro-lateral view, c, left side of pelvis of NMQR 2902 in lateral view. Hatching
indicates a break through the bone. Scale represents 10 mm.

acetabulum and dorsal blade. This gives the bone a tall, thin, somewhat delicate appearance. At least two and
possibly three ribs make contact with the ilium, compared with five in D. trigonocephalus , four in Kingoria and
three in Patranomodon and Cistecephalus. The most posterior of these ribs would have attached ventro-
posteriorly on the neck of the ilium. The anterior and posterior corners of the acetabulum are rounded and
built up where the ischium and pubis articulate. The acetabulum (Text-fig. 5c) is roughly hemispherical.

Pubo-ischiadic plate

The pubis (Text-fig. 5b; pu.) is a thin L-shaped bone, the short limb of the L making contact with the antero-
ventral corner of the acetabulum, and the long limb being directed postero-ventrally to meet the ischium. There
is a large pubo-ischiadic fenestra (Text-fig. 5b; pu.isc.f.) between the two bones. The ischium is basically a
square bone from which a large notch has been cut antero-ventrally to form one half of the pubo-ischiadic
fenestra. Both bones are thin plates except where they are built up proximally to form their contributions to
the acetabulum.

Femur

The femur (Text-fig. 6a) is a fairly slender bone, c. 10 per cent longer than the humerus. The head (Text-fig.
6a; he.) has a reasonably distinct condyle which is offset from the proximo-anterior corner of the bone, giving
the femur a very slight S-curvature. The distal condyles are not well marked, in contrast to those of the
humerus. The trochanter major (Text-fig. 6a; tr.m.) is not large, and is not set off from the shaft of the bone
as in D. trigonocephalus , but grades into the shaft as in Robertia. However, the femur of Eodicynodon appears
to be more slender than that of Robertia , partly due to less expansion of the proximal and distal ends (especially
the proximal) and partly because the shaft portion of the bone is proportionally longer.

Tibia and fibula

These are both slender bones (Text-fig. 6b-c), c. 80 per cent the length of the femur. There is a well-marked
cnemial crest on the tibia, but little to distinguish them from these same elements in other genera.

RUBIDGE ET AL.. EODICYNODON POSTCRANIUM

403

text-fig. 6. Femur, tibia and fibula of Eodicynodon NMQR 3153. a, right femur in anterior view, b, left tibia
in anterior view, c, left fibula in anterior view. Scale represents 10 mm.

Pes

Little information is available concerning the structure of the pes. Ungual phalanges are slender and
approximately the same length as other phalanges. The calcaneum is a flat ovoid plate of bone. The astragalus
is more spherical and a deep notch can be seen in the surface that is visible. These elements appear to be very
similar to those described in Kingoria. Few other tarsi are available for comparison.

Vertebrae and ribs

The axial skeleton (including the atlas-axis complex) does not appear to have any features which would
distinguish Eodicynodon from other dicynodont genera. No specimen is sufficiently complete to permit an
estimate of the number of presacral vertebrae. There are probably three sacral vertebrae. At least six caudal
vertebrae are present.

DISTINCTIVE FEATURES OF EODICYNODON POSTCRANI AL ELEMENTS

This description permits several features of Eodicynodon to be recognized which distinguish this
genus from all other dicynodont genera: (i) the scapula has a broad dorsal edge and is not so
constricted above the acromion process as in other genera; (ii) the coracoid plate is large in

404

PALAEONTOLOGY, VOLUME 37

comparison with the scapula; (iii) the humerus has a large plate-like triangular delto-pectoral crest,
well-developed distal condyles, and proximal and distal ends at a large angle to one another; (iv)
the clavicles have very large expanded ends; (v) the radius and ulna are slender bones. In the pelvic
girdle: (vi) the pubo-ischiadic plate is very large; (vii) the ilium is tall and thin with a small anterior
process; (viii) the femur is a slender bone with a long shaft and comparatively little expansion of
the proximal end.

Owing to the shape of the delto-pectoral crest, the humerus is the most distinctive bone of the
postcranium and would serve alone to distinguish Eodicynodon from other genera. The proportions
of the ilium are also different from those of other dicynodont ilia.

Recently it has become apparent that for stratigraphical purposes an important distinction is to
be able to differentiate Eodicynodon from Robertia. Currently the stratigraphical range of Robertia
is uncertain, largely because of the lack of an adequate generic diagnosis of this form. The type of
Robertia has been more fully prepared and described (King and Rubidge 1993) and it is now evident
that there are many more specimens of Robertia in collections than were previously recognized.
Systematic fossil collecting in the lowermost rocks of the Beaufort Group has revealed that
Eodicynodon and Robertia occur together in the upper parts of the Eodicynodon-Tapinocaninus
assemblage zone (Rubidge 199(G).

The humerus most easily distinguishes Eodicynodon from Robertia. The distinction is not based
on bone element proportions, and therefore comparative material (either from the rest of the
skeleton, or from other genera) is not necessary. However, the ilium, femur, scapula, and coracoid
plate all show differences in proportions from Robertia as indicated above, and comparative
material would distinguish between the two genera.

FUNCTIONAL MORPHOLOGY

Eodicynodon has a distinctive postcranial morphology. In some respects this morphology appears
to comprise less derived character states than other dicynodonts: the large ventral parts of the limb
girdles, the robust humerus with its expanded ends, the very small iliac blade with small anterior
process, the small number of sacral ribs. However, in other respects Eodicynodon is as derived as
later dicynodonts: the well-developed and everted acromion process, and the absence of any
remnant of the ectepicondylar foramen in the humerus.

It is clear, that as is usual for most dicynodonts, there is a large difference in morphology between
the fore- and hindlimbs. Although the hindlimb is the longer, it is much more gracile. However, the
discrepancy in Eodicynodon is more marked than in other less derived dicynodonts such as
Dicynodon trigonocephalus : the ends of the humerus are more expanded, and the distal condyles are
very prominent. These features, together with the robust ventral shoulder girdle, slightly expanded
dorsal margin of the scapula, large olecranon process of the ulna, and broad manus with long sharp
claws are reminiscent of those found in digging forms such as Cistecephalus (Cluver 1978) and
Kawingasaurus (Cox 1972) which used forearm flexion to dig, although the features are much less
marked in Eodicynodon. However, it is much more likely that these features reflect the primitive
condition of the forelimb and girdle in dicynodonts.

In the pelycosaur Dimetrodon the scapula also flares dorsally (Romer 1922). The posterior,
anterior and dorsal margins of the scapula blade, which are emphasized by this flaring, are where
postural muscles such as the levator scapulae and serratus anterior inserted. It is possible that
narrowing of the blade in later dicynodonts and therapsids in general is correlated with a reduction
in this part of the postural musculature as a more ‘mammalian’ stance was adopted. However, the
reduction of postural musculature does not seem to be a simple process in dicynodonts. Although
it seems that the superficial postural muscles were reduced, other muscles involved in support of the
anterior part of the body (the deltoids, pectoralis and biceps) were emphasized. This emphasis is
reflected in the form of the delto-pectoral crest.

The distinctive appearance of the delto-pectoral crest of Eodicynodon compared with that of other
dicynodonts is due to a lack of development of the distal part of the crest (Text-fig. 7). Muscles

RUBIDGE ET AL : EODICYNODON POSTCRANIUM

405

text-fig. 7. Humerus of Eodicynodon
superimposed on that of Dicynodon
showing the rotation of the distal end
relative to the proximal (arrow), and the
larger development of the distal part of
the delto-pectoral crest. The proximal
end is in dorsal view.

he.

which would have attached to this area of the crest in dicynodonts were part of the deltoideus, part
of the pectoralis, and the biceps and brachialis (King 1 98 la). Developing the delto-pectoral crest
distally, as happened in more derived dicynodonts, allowed the deltoideus and pectoralis to insert
more distally. Being farther from the articulation of the humerus, these muscles would have acted
more powerfully (but also more slowly) on the humerus. Elevation and protraction (deltoideus) and
depression and protraction (pectoralis) would have become powerful actions. Being farther from the
articulation also made the muscles more effective in a postural role (Text-fig. 8).

The drawback of this arrangement is that the biceps muscle would have become rather short and
its role in limb flexion and support was presumably diminished.

The emphasis on the activity of the deltoideus and pectoralis muscles in dicynodonts more
derived than Eodicynodon may be reflecting in these forms an emphasis on powerful, if slow.

text-fig. 8. Schematic representation of the deltoid and pectoral musculature of a, Dicynodon and
b, Eodicynodon. The diagrams show a cross-section through the body, head towards the observer.

Scale - 10 mm.

406

PALAEONTOLOGY. VOLUME 37

movement and possibly larger body size. This corroborates the usual interpretation of most
dicynodonts as slow but powerful animals. Powerful anterior postural musculature may have been
important to dicynodonts in supporting a large abdomen in which plant matter was slowly digested,
and in supporting the robust head equipped with powerful jaw adductor muscles.

The new insertional position of the deltoideus and pectoralis may be one of the reasons that the
coracoid-procoracoid plate, so extensive in Eodicynodon, has been reduced in later forms. The more
advantageously placed pectoralis muscle might thus have required a less extensive area of origin.

The greater angulation of the proximal and distal ends of the humerus reflects a condition where
the humerus projects laterally from the glenoid. If the humerus is positioned so that the distal end
of the bone is horizontal, the delto-pectoral crest hangs down as a near-vertical plate. So positioned,
there is no restriction on the large size of the plate which can accommodate large postural, as well
as limb-protracting, muscles. The vertical alignment of the delto-pectoral crest also allows muscles
to pass under the smooth antero-ventral edge of the shaft of the bone to attach to the ventral
ectepicondylar surface.

The origins of the supracoracoideus and scapulo-humeralis in Eodicynodon show an intermediate
condition between those of pelycosaurs and those of later dicynodonts. The coracoid plate is
extremely large and no doubt accommodated a large part of the origin. However, a strongly everted
acromion process is also present. The lower edge of this is sharp and does not provide a route for
the supracoracoideus to reach the medial side of the scapula. There is, however, a smooth hollow
ventral to the acromion process where part of the supracoracoideus might have originated. This
would confirm the suggestion (King 1981/ff that the everted acromion process had initially very little
to do with muscle attachment.

The primitive nature of the forelimb and pectoral girdle is reinforced by the pelvic girdle and limb
where the small anterior process of the ilium and the lack of development of the trochanter major
would also be primitive features, reflecting the lack of a forwardly-pulling part of the ilio-femoralis,
the precursor of the mammalian gluteals. The height of the ilium would also enhance the levator
action of the ilio-femoralis. Presumably at the Eodicynodon stage the femur was protracted mainly
by the pubo-ischio-femoralis interims. Retraction would have been accomplished by part of the ilio-
femoralis on the posterior iliac process, the ischio-trochantericus and the caudi-femoralis. The large
ischiadic plate would have accommodated adequate ischio-trochantericus muscles. As far as the
caudi-femoralis is concerned, in Eodicynodon there are at least four post-sacral vertebrae which have
long, flat, robust ribs attached to them which would have afforded an origin for the caudi-femoralis.
However there is no obvious insertion for the muscle on the femur.

Acknowledgements . We are very grateful to Dr C. M. Engelbrecht and Mr J. Welman (National Museum,
Bloemfontein) and Mr R. Oosthuizen for permission to study material in their care, and to Mr J. Nyaphuli,
Mr Z. Gregorowski, Ms D. Jaeger and Mr J. Motho for their skilful preparation of the material. Dr Saskia
Waters produced the illustrations.

REFERENCES

barry, T. H. 1974. A new dicynodont ancestor from the Upper Ecca. Annals of the South African Museum , 64,
117-136.

brinkman, d. 1981. The structure and relationships of the dromasaurs (Reptilia: Therapsida). Breviora , 465,
1-34.

cluver, m. a. 1978. The skeleton of the mammal-like reptile Cistecephalus with evidence for a fossorial mode
of life. Annals of the South A frican Museum , 76, 213-246.

and king, G. m. 1983. A reassessment of the relationships of Permian Dicynodontia (Reptilia, Therapsida)
and a new classification of dicynodonts. Annals of the South African Museum , 91, 195-273.

RUBIDGE ET AL.\ EODICYNODON POSTCRANIUM

407

cox, c. b. 1959. On the anatomy of a new dicynodont genus with evidence of the position of the tympanum.
Proceedings of the Zoological Society of London , 132, 321-367.

— 1972. A new digging dicynodont from the Upper Permian of Tanzania. 173-190. In joysey, k. a. and
kemp, t. s. (eds). Studies in vertebrate evolution. Oliver and Boyd, Edinburgh, 284 pp.

king, G. M. 1981a. The functional anatomy of a Permian dicynodont. Philosophical Transactions of the Royal
Society , Series B , 291, 243-322.

— 19816. The postcranial skeleton of Robertia broomiana , an early dicynodont (Reptilia, Therapsida) from
the South African Karoo. Annals of the South African Museum , 84, 203-231.

— 1985. The postcranial skeleton of Kingoria nowacki (von Huene) (Therapsida; Dicynodontia). Zoological
Journal of the Linnean Society, 84, 263-289.

1988. Anomodontia. Encyclopedia of paleoherpetology, 17C, 1 1 74.

— and rubidge, b. s. 1993. A taxonomic revision of small dicynodonts with postcanine teeth. Zoological
Journal of the Linnean Society, 107, 131-154.

— oelofsen, b. w. and rubidge, b. s. 1989. The evolution of the dicynodont feeding system. Zoological
Journal of the Linnean Society, 96, 185-21 I.

romer, a. s. 1922. The locomotor apparatus of certain primitive and mammal-like reptiles. Bulletin of the
American Museum of Natural History, 46, 5 1 7-606.

rubidge, b. s. 1984. The cranial morphology and palaeoenvironment of Eodicynodon Barry (Therapsida:
Dicynodontia). Navorsinge van die Nasionale Museum Bloemfontein , 4, 325-402.

— 1990m A new vertebrate biozone at the base of the Beaufort Group, Karoo Sequence (South Africa).
Palaeontologia Africana, 27, 17-20.

— 19906. Redescription of the cranial morphology of Eodicynodon oosthuizeni (Therapsida: Dicynodontia).
Navorsinge van die Nasionale Museum Bloemfontein, 7, 1-25.

— 1990c. The cranial morphology of a new species of the genus Eodicynodon (Therapsida: Dicynodontia).
Navorsinge van die Nasionale Museum Bloemfontein , 7, 29 — 42.

— and hopson, j. a 1990. A new anomodont therapsid from South Africa and its bearing on the ancestry
of Dicynodontia. South African Journal of Science, 86, 43-45.

S. RUBIDGE

Bernard Price Institute for Palaeontological Research
University of the Witwatersrand
Private Bag 3

P.O. Wits 2050, South Africa

G. M. KING

South African Museum
P.O. Box 61

Cape Town 8000, South Africa

p. J. HANCOX

Typescript received 24 May 1993
Revised typescript received 4 October 1993

Bernard Price Institute for Palaeontological Research
University of the Witwatersrand
Private Bag 3

P.O. Wits 2050, South Africa

408

PALAEONTOLOGY, VOLUME 37

ABBREVIATIONS

ac.p.

acromion process

Eod.

Eodicynodon

pec.

pectoralis muscle

ant.p.

anterior process

Eod. d.p.cr.

Eodicynodon delto-
pectoral crest

pu.

pubis

bic.

biceps muscle

fl.

attachment area for
flexor muscles

pu.isc.f.

pubo-ischiadic fenestr

cor.

coracoid

gl-

glenoid

ra.c.

radial condyle

d.p.cr.

delto-pectoral crest

he.

head

scap.

scapula

del.

deltoideus muscle

hum.

humerus

tr.m.

trochanter major

Dicy.

Dicynodon

il.

ilium

lll.C.

ulnar condyle

Dicy. d.p.cr.

Dicynodon delto-
pectoral crest

isc.

pc.

pc.r.

ischium
procoracoid
procoracoid foramen

I-V

digits 1 to V

A COMPUTER MODEL FOR SKELETAL GROWTH
OF STROM ATOPOROIDS

by ANDREW R. H. SWAN and STEPHEN KERSHAW

Abstract. A robust and versatile computer model of simple accretionary laminar growth can be developed
based on probabilistic accretion of pixels on a raster array. The model is a reasonable analogue for growth of
simple organisms such as stromatoporoids. Experiments with the model allow the effects of sedimentation and
various alternative growth algorithms to be simulated. The model can be validated, with some reservations, on
theoretical and empirical bases: the simulations show similarities to observed stromatoporoid morphologies.
The results suggest that morphology is strongly influenced by the pattern of sedimentation and that
stromatoporoids required a local pause in sedimentation in order to become established. The results are
consistent with the view that stromatoporoids were integrated organisms but with a low level of organization
allowing a degree of autonomy of modular growth.

Fossil stromatoporoids were sponges which secreted a secondary calcareous skeleton within a soft
tissue coating. Similar modern sponges reveal that soft tissue is limited to the upper few millimetres
of the skeleton; as the sponge grows, the underlying skeleton is vacated (Hartman and Goreau
1970; Stearn 1972, 1975). The final result is a laminated skeleton which can display a range of
growth geometries. This study presents a computer-based model to simulate skeletal growth in
stromatoporoids, and demonstrates significant parallels between real and simulated forms. The
model allows exploration of the controls on growth of stromatoporoids and assists palaeoecological
analysis of these fossils.

The term ‘stromatoporoid' refers to a particular organization of skeletal structure, typically
comprising sheet-like lateral elements (laminae) and vertical rod-like elements (pillars) arranged as
a reticulum. Stromatoporoids are traditionally regarded as a taxonomic unit within the Porifera
(Stearn 1975), but more recent views are that the stromatoporoid structure represents a grade of
organization of sponge skeletons unrelated to taxonomy (Vacelet 1985); Wood (1990) recognized
four grades in sponges, with some degree of overlap.

These conflicting views of the status of stromatoporoid structure are not relevant to the
morphological computer simulations presented here, but two skeletal structures amongst sponges
generate remarkably similar external shapes: stromatoporoid and chaetetid. Chaetetids differ from
stromatoporoids in being composed of small tubes. In both cases, the morphology is well
documented and both exist as skeletons containing growth lines, so that the skeleton can be
considered as an accreted pile of laminar units. The shape of successive laminae may change through
growth, such that the overall result may vary from a flat sheet to a column. Sponges are modular
rather than colonial organisms, and therefore it is appropriate to refer to each sponge as an
individual rather than a colony. Coloniality in sponges is therefore truly at the cellular level. Wood
et al. (1992), drawing on work by Jackson (1983) and others, discussed the relationship between
modularity and growth success in sponges, and concluded that highly modular organisms live long
lives and grow to large sizes, and can therefore be successful at activities such as reef building.

Earlier works use the term coenosteum to describe a single stromatoporoid; coenosteum derives
from hydrozoan coelenterate terminology, from the days when stromatoporoids were generally
considered as hydrozoa. The term coenosteum should now be abandoned in stromatoporoid
terminology (Stearn 1984, p. 316).

| Palaeontology, Vol. 37, Part 2, 1994, pp. 409-423|

© The Palaeontological Association

410

PALAEONTOLOGY, VOLUME 37

There have been a number of attempts at understanding the morphology of stromatoporoids
and related organisms. Kershaw and Riding (1978, 1980) illustrated field geometries of most
stromatoporoids using triangular arrays, and a similar approach was applied to corals by Young
and Scrutton (1991). Kershaw and West (1991) discussed the geometry of laminar growth for
chaetetids and noted that such an approach could be applied to other calcareous skeletal organisms,
such as trepostome bryozoans (Ross 1987) and tabulate corals. The use of computer simulation is
a new initiative in this field.

COMPUTER MODELLING OF SKELETAL GROWTH

Computer models are often based on equations which describe the processes in the system: the
equations may be derived empirically or from theoretical considerations. This is the approach used,
for example, to model growth of molluscan shells by Raup (1966), of stromatoporoids by Hofmann
(1969), of corals by Gratis and Macintyre (1976), and to model accumulation of reef carbonates by
Bosence and Waltham (1990). Such mathematical models could, in principle, be devised for
accretionary systems. To cite mundane examples, the formation of hailstones (or lapilli) by
accretion in a cloud, or of ooliths in agitated shallow seas, could be modelled by the equation for
a sphere with successively increasing radii, and this could be extended to constructing half-spheres
to emulate accretion around a nucleus on the sea bed. This mathematical modelling approach has
been rejected for the present context for two main reasons. Firstly, any equation that might be used
could not be regarded as in any way inherent in the organism; it would not emulate genetic
programming (blue-green algae do not make ooliths because they are genetically programmed with
the equation of a sphere). Secondly, apart from very simple situations, the dynamic and
probabilistic aspects of growth become too complex for mathematical models.

An alternative approach is to devise simple rules which can be applied to any point on the
simulated organism. These can be sufficiently simple and general that they can readily be regarded
as a result of genetic control, basic physiological response or external physical effects. In practice,
it is convenient to construct the simulated organism on a grid or raster, which can be directly
represented on the pixel array of the computer screen. Consequently, the rules for accretion are
designed to be applicable to each pixel in the raster array. This approach allows each small part of
the simulated organism to operate autonomously, with no central control imposed by the organism
as an integral unit. This is in contrast to the mathematical modelling approach, in which exact
control of all parts of the organism is implied by the use of an equation. Indeed, success at modelling
an organism using the raster approach may have implications about the level of organisation of the
organism.

DEVELOPMENT OF A MODEL FOR STROMATOPOROID GROWTH
Accretion on a raster array

Accretionary growth simply involves the addition of new material on existing surfaces; usually
surfaces that have resulted from previous increments of accretion. In terms of the computer model,
accretion involves ‘switching on’ only these pixels which are immediately adjacent to already
‘switched on’ pixels. However, although the raster array is a convenient way of accounting for
occupation of space, it is unnatural in that; (1) each raster cell or pixel is a quantum which bears
no relation in size, shape or position to the units of accretion in the real organism (which may be
biological cells or even molecules): it is computationally awkward to account for fractions of pixels;
(2) the raster imposes an artificial square anisotropy on the model.

The effect of these problems can be demonstrated by considering accretion around a simple one
cell nucleus (Text-fig. 1a). If all neighbouring cells are ‘switched on’, over several increments we

SWAN AND KERSHAW: STROMATOPOROID GROWTH

411

text-fig. 1 . Accretion on a square array or raster of
pixels, a, deterministic accretion, producing a result
that reflects the square anisotropy of the raster, b,
stochastic accretion, where the accretion of a pixel is
never certain; probabilities can be fixed so that the
structure accretes at the same rate in all directions.

A

accumulate a structure with square symmetry, rather like a crystal but not like organic accretionary
structures. This is because accretion of one cell in the direction of the 45° diagonals requires
accretion of one horizontal and one vertical increment, so these directions are more slowly extended.
The square anisotropy creates the problem, and the idea of pixels as quanta prevents the solution
of the problem by use of fractional pixel accretion (e.g. fill two-thirds of a diagonal pixel for each
horizontal or vertical increment).

The problem can be solved by probabilistic accretion. In this scheme, pixels are never certain to
be accreted to the structure; we attach probabilities of accretion to all candidate pixels, where the
probabilities are determined by various aspects of spatial location. This, as we shall demonstrate,
is the crucial aspect that confers great flexibility and robustness to the model. We can devise rules
for determining the probabilities to be attached to candidate pixels on the basis of spatial
disposition, such that these are analogous to the varying growth rates of different parts of a real
organism.

The problem with the 45° diagonal pixels is that 2 increments of purely horizontal or vertical
accretion extend the structure 2 units in those directions, but diagonal accretion in 2 increments
(1 horizontal and I vertical) extend the structure only \J2 units. So, the rate of diagonal accretion
needs to be enhanced relative to the vertical or horizontal rates. In a probabilistic model, this can
be done by increasing the likelihood of ‘switching on’ a pixel if it has two rather than one ‘switched
on’ neighbours (Text-fig. 1b).

Description of basic model

The initial state of the basic model is a flat sediment surface and a single pixel ‘seed’ for the
structure. Pixels not occupied by the simulated organism are coded zero; the structure is made of
pixels with non-zero pixel colour codes. In each increment, adjacent pixels are scanned and are
allocated probabilities of becoming non-zero (‘switched on’) according to position.

412

PALAEONTOLOGY, VOLUME 37

If there are no non-zero pixels amongst the four immediate neighbours, or if the position is
occupied by sediment, the probability P is zero. If there are n non-zero pixels amongst the four
immediate neighbours and the position is not occupied by sediment, the probability P is:

The status of a pixel is decided as follows: a random number is generated in the range 0 to 1 ; if the
random number does not exceed P, the pixel becomes non-zero; otherwise it remains ‘off’. In
practice, this algorithm is found to create good approximations to circles and semicircles in simple
accreting systems. The result of the simplest type of run of the algorithm is shown in Text-figure 2.

text-fig. 2. Simulation resulting from the basic model
for accretion around a single-pixel nucleus on a
sediment surface. In this and subsequent simulations,
there are about 4 pixels per millimetre.

In order to show the shape of successive increments of accretion, the non-zero pixels are allocated
alternating colour codes (black and stippled on hard-copy) after every fixed number of iterations
(eight in the simulations presented here).

The basic model can be adapted to incorporate the effects of various factors by incorporating
more complex rules for calculating pixel probabilities.

Geotropism

Geotropism is a very common attribute of organisms; negative geotropism (preferentially growing
upwards) is likely to be useful for moving clear of turbid sediment or competition, and closer to
light. This is difficult to distinguish from positive phototropism (see Graus and Macintyre 1976),
which may have the same purpose and effect; these are not distinguished in the versions of the model
presented here. Geotropism is incorporated into the model by introducing a different probability-
weighting to pixels with vertically disposed non-zero neighbours. A geotropic factor 'geo' is
specified by the user. If the candidate pixel is immediately above a non-zero pixel and has no other
non-zero neighbours, its probability of being accredited is:

geo
geo + 1

(2)

(Compare with equation 1). If the candidate pixel is immediately lateral to a non-zero pixel and has
no other non-zero neighbours, its probability of being accreted is:

P =

\jgeo

( 1 /geo) + 1 '

(3)

Consequently, the probability is enhanced if there is a non-zero pixel below, but reduced for every
lateral non-zero pixel. Values of geo greater than 1 result in more rapid vertical growth; geo less
than 1 inhibits vertical growth (Text-fig. 3).

Effect of sedimentation

As noted above, no pixel is accreted onto the structure if it is in a position occupied by sediment.
This has a mundane consequence if the sediment surface remains constant, but this is unrealistic.
Stromatoporoids are at least sometimes associated with significant sedimentation rates and we

SWAN AND KERSHAW: STROM ATOPOROI D GROWTH

413

text-fig. 3. Simple accretionary structures with varying value of the geotropism factor geo.

expect sediment accumulation to affect the growth of the structure. In the computer model, only
those pixels above the sediment surface are scanned as candidates for accretion, so a vertical
addition to the surface position will automatically simulate smothering of growth sites by sediment
cover. The sediment surface could be controlled manually after each iteration of the program, but
experiments with the model have focused on the alteration of three sedimentation parameters:

(1) interval between sediment increments (units: number of iterations of accretion algorithm);

(2) amount of sediment in each increment (units: pixels); and (3) length of hiatus during initial
establishment of structure (units: number of iterations of accretion algorithm).

Experimental simulations can be run to demonstrate the effect of each parameter. This is best
shown by means of an array of simulations on a 2-dimensional space defined by: (1) interval
between sediment increments; and (2) ratio of amount of sediment in each increment to the interval
between sediment increments. Such an array is shown for the simple accretionary model in Text-
figure 4 and demonstrates the following properties.

(1 ) The partial occlusion of the accreting surface by sedimentation, followed by growth back over
the sediment surface during quiescent intervals, results (predictably) in a ragged lateral edge to the
structure. This property was quantified using a raggedness index by Kershaw (1984).

(2) Where sedimentation is significant and consistent, the bases of structures have a clear conical
geometry.

(3) Once established in a regime of regular sedimentation, structures grow indefinitely.

(4) The occupancy of the 2-dimensional parameter space, and hence the diversity of geometries,
is constrained by the condition where sedimentation rate exceeds growth rate.

As we shall see, points (2), (3) and (4) are not compatible with observations of real
stromatoporoids. The range of possibilities can, however, be extended by using the third parameter
cited above (allowing a hiatus in sedimentation). Using a hiatus of 100 units, a comparable array
of forms can be generated (Text-fig. 5), including high-domed morphologies with on-lap of laminae.

414 PALAEONTOLOGY, VOLUME 37

log(interval)

text-fig. 4. Array of simulations with varying interval between sediment increments and amount of sediment
in each increment. Horizontal dotted lines show successive positions of sediment surface. The forms shown at
the top of the array represent the highest values of relative sedimentation rate at which structures can develop.
Above this, the fundamental constraint of average sedimentation rate exceeding average growth rate applies;
in this region the structures are extinguished soon after initiation. Also, note the prevalence of conical bases

and ragged edges.

Response to sediment surface

The basic model includes rules which regard the sediment surface as neutral; probabilities attached
to pixels adjacent to sediment are not enhanced or reduced. However, it is conceivable that
proximity to sediment could either inhibit growth (see section on competition below) or enhance
growth. The positive effect is introduced by a modification of the basic model, allowing sediment
to have a similar effect as the pre-existing accreted structure in influencing the probability of
inclusion of new pixels. The probability calculation P — n/(n+ 1) (equation 1) is applied with the
specification that n is the number of non-zero adjacent pixels, including those with sediment colour
codes, but with the proviso that one of the n must belong to the accreting structure. This assures

SWAN AND KERSHAW: STROM ATOPORO I D GROWTH

415

-0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

log(interval)

text-fig. 5. Array of simulations similar to that shown in Text-figure 4, except that structures are permitted
to become established during a hiatus of 100 units. Most structures are eventually over-run by sedimentation,
but this array includes different geometries, typically with on-lap of laminae. Geometries with lower relative
sedimentation rates are possible; these are similar to those in Text-figure 4, but broader.

that structures do not nucleate arbitrarily across the whole sediment surface. There is an analogous
modification to equation 2 if the geo factor is used. These modifications produce structures with
broader bases (Text-fig. 6a), which may be advantageous for stability and for excluding spatial
competitors.

416

PALAEONTOLOGY, VOLUME 37

text-fig, 6. a, structure produced by modified model in which contact with sediment has a positive effect on
accretion, b. merging of structures modelled by application of the basic model to an initial state where there

are two ‘seeds’.

Merging of multiple colonies

The algorithms described above can be equally well applied to initial states in which there are two
or more ‘seeds’. In the simple versions of the program, there is no sense of individual identity of
pixels or structures, so separate structures merge together. Further increments have the effect of
masking the original separate structures (Text-fig. 6b).

Competition

The simple algorithms so far described treat pixels equally, regardless of general position on the
developing structure. This could be regarded as unrealistic; for example, positions projecting on the
top of the structure may be better favoured for access to food or oxygen, when compared with
positions in crevices or adjacent to the sediment. This may have a number of different types of result,
depending on the level of organization of the organism.

(1) If local parts of the structure are highly autonomous, they will compete with each other and
favoured parts of the structure may grow at the expense of others.

(2) If resources are to some extent distributed, growth may be equal regardless of site.

(3) If growth is highly centrally organized, there may be strategies of: (a) enhanced growth at
favoured locations to take advantage of resources; or (b) systematic patterns of growth determined
by other criteria (e.g. structural strength). There may be genetic control of growth form underlying
such strategies (Kershaw 1990).

The basic model described in previous sections is based on type 2. A model for type 1 will now
be described, the results of which may be indistinguishable from those of type 3 a. Type 3 b is,
however, beyond the scope of the current suite of models.

If pixels are autonomous, we can model a situation in which pixels are more likely to form sites
of accretion if they are in favoured positions. Positive feedback in growth can be achieved if pixels
in more open projecting positions are favoured over those in enclosed positions; this is intuitively
reasonable. The ‘openness’ of the position is assessed by the modified program by counting the
number of non-zero pixels from a circular scan of 16 points at a given radius from the candidate
pixel and at 22-5° intervals. The number of the 16 that are occupied by the structure or sediment

text-fig. 7. Stages of growth of a structure using a model with local autonomy and competition. Accretion
is enhanced at more open sites and inhibited at enclosed sites.

SWAN AND KERSHAW: STROM ATOPO RO I D GROWTH

417

may be symbolized by /?; the revised probability to be attached to the candidate pixel P' is then
calculated from the probability P (from equations 1 3) by:

F = PR,S. (4)

Consequently, positions on flat surfaces (R = 8) have unaffected neutral growth rates but open sites
(R < 8) give increased growth (higher P) and enclosed sites (R > 8) are inhibited (lower P). The
result of this algorithm (Text-fig. 7) is to enhance arbitrary local projections into major lobate
branches; this can be allowed to develop into a dendritic structure. The width of the branches is
determined by the specified scan radius, representing the range around a point on the organism
within which the spatial arrangement of the structure has some influence.

Competition between multiple colonies

If the ‘competition’ algorithm described above is applied to multiply-seeded structures, they do not
merge. Growth is inhibited when it takes a structure to within the specified radius of a rival
structure. Smooth sided structures result (Text-fig. 8). In the case of local autonomy of growth, this
result should be associated with lobate branching elsewhere on the structure. In the case of
distributed growth (types 2 and 3 above), these two effects will not necessarily be associated.

text-fig. 8. Simulation of effect of two structures on
each other when growth is inhibited at enclosed sites.

VALIDATION

The validation of computer models is always problematical. It is based on two principles:
(1) assessment of processes and parameters; (2) comparison of simulations with reality.

Processes and parameters

The process involved here is accretionary growth; the parameters are the variables which control
the rates and sites of growth. Many computer models are able to incorporate calibration of
parameter values by using empirical data from recent analogues : thus Bosence and Waltham ( 1 990)
were able to include real rates of growth in their coral reef model. In the present study, there are
no reliable sources for such information; attempts to relate growth banding to growth rate have
required basic assumptions, such as Meyer’s (1981) study where annual increments were assumed.
Calibration on this basis can be done but does not provide unequivocal data on growth rates. We
can, nevertheless, appraise the pertinence of the processes in the model in relation to apparent
growth mechanisms of fossil organisms.

The raster accretion method of this study treats a skeleton as composed of minute, equal-sized
units (pixels), and therefore ignores the complexities of skeletal differentiation between the various
groups of taxa with similar gross morphotypes. Its application to growth of organisms with

418

PALAEONTOLOGY, VOLUME 37

relatively low modular integration, such as favositid corals, may not be so satisfactory, but when
applied to highly integrated skeletons such as many sponges, the model may be closer to reality.

Laminar accretionary growth provides a useful concept for visualizing the overall geometric
development of a skeleton of these organisms, but does not provide a means for modelling the
growth mechanism. Such growth assumes that accretion took place simultaneously across the
skeleton surface, but examples of stromatoporoids occur for which this was demonstrably not the
case, so this concept is only partially applicable. Also, laminar growth units are recognizable as
geometric entities in a wide range of phyla with different skeletal organizations, and are a reflection
of the need for growth. However, they give no information about intrinsic controls in individual
phyla. Laminar growth occurs in organisms with obvious differences in skeletal organization in
clonal organisms (Jackson 1983), and especially in terms of the degree of integration, or
modularity — a crucial concept in studies of clonal organisms. In tabulate corals such as Favosites
a module is clearly identifiable as a single corallite, whereas in sponges modules are not so easily
recognized because there are no identifiable individuals. However, because sponges filter-feed, they
consist of tissue arranged in an incurrent-excurrent system. In stromatoporoids, the centres of
excurrent flow are astrorhizae (sets of branching root-like grooves on the upper surface of many
species). These are often arranged in an evenly-spaced pattern on stromatoporoid surfaces, so that
water is drawn in through the tissue around the astrorhizae, and waste water is channelled to the
astrorhizal centre and expelled (LaBarbera and Boyajian 1991). Astrorhizae therefore provide
evidence of aquiferous units with unclear boundaries, which could be regarded as the closest
approximation to individuals in sponges. Apart from this, sponges only show individuality at the
cellular level, not recognizable in fossils. Unfortunately, not all stromatoporoids show astrorhizae,
so the aquiferous limit is not a universally quantifiable feature. Using a modular approach, it is clear
that in the wide variety of organisms with laminar accretionary growth, modules have different sizes
and types. Modelling the growth of such a disparate array of skeletal constructions may therefore
require variety of approach.

Stromatoporoid skeletal structure varies from organizations with prominent laminae to those
with prominent pillars, and as a result there are no definable subunits of growth which can be
recognized in all stromatoporoids, unlike tabulates or even chaetetids which have tubes as the
smallest unit of growth. Consequently, growth was presumably quite locally organized in
stromatoporoids (Wood 1991). Therefore the raster approach adopted here appears to be closer to
the way stromatoporoids grew than for the other groups.

Within the sponges, stromatoporoids and chaetetids do not show uniform growth. Kershaw and
West (1991) showed considerable internal complexity in calicle distribution in chaetetids within
single individuals. In stromatoporoids, variation of internal skeletal elements occurs where parts of
an individual display prominent laminae while others show prominent pillars, and some
stromatoporoids show phases of growth (Stearn 1989). Stromatoporoids may show these variations
on a rhythmic basis, which suggests an environmental control on the growth of successive layers of
skeleton. Modelling using the raster approach adopted here is unlikely to resolve such fine scale
variation. Furthermore, the stochastic nature of the computer model has no analogue in growth of
real organisms : it is used as a convenient means of emulating curved increments of accretion on the
raster array of the computer device. It is clear that the application here is an algorithmic
approximation of real growth. Thus it does not explain how growth occurs, but is a proxy for
demonstrating the geometry of growth in a skeleton. However, the basic processes of the model such
as accretion on ‘live’ surfaces and smothering by sediment are highly plausible as properties of real
organisms and have crucial influence on the final geometry of the structure.

Comparison with real forms

The simulations presented in Text-figures 4 and 5 demonstrate the dependence of shape on
sedimentation. This is shown, for convenience, only for cases of regular sedimentation increments
and intervals. The morphology of real forms will be determined by the effectively arbitrary history

SWAN AND KERSHAW: STROM ATOPOROI D GROWTH

419

of sedimentation and other real events. Furthermore, the shape of the initial substrate often differs
from the flat geometry of the simulations. Consequently, we do not expect an exact fit of a
simulation to any particular stromatoporoid specimen; rather, we should compare types and ranges
of geometry.

Correspondences. There appears to be a correspondence between simulations and real forms in the
following aspects.

(1) General morphology. The simulations generally produce types of massive domed structure,
corresponding to the typical stromatoporoid morphology. This is not entirely a mundane
observation ; it is an important aspect of stromatoporoid morphology that more complex structures
are not as typical as they are of other groups, such as corals. The special modifications that can be
made to the algorithm to produce more complex structures may be analogous to the more complex
growth strategies of other organisms and atypical stromatoporoids.

(2) Geotropism. The range of geometries resulting from changes to the geotropism factor (Text-
fig. 3) match the range of degrees of convexity of real stromatoporoids (Text-fig. 9), described by

text-fig. 9. Examples of contrasting stromatoporoid geometries that can be modelled by varying values of the
geotropism factor (see Text-figure 3). Notice also the concave bases, suggesting establishment on convex local
highs on the substrate. All are traced from photographs of specimens from Silurian of Gotland. Scale bars

represent 10mm.

the continuum from laminar to low domical to extended domical morphotypes by Kershaw and
Riding (1978).

(3) Ragged edges. The ragged lateral margins resulting from simulations involving periodic
sedimentation (Text-figs 4-5) have the same form and the same inferred cause as those in real
stromatoporoids and chaetetids (Text-fig. 10), as documented by Kershaw and Riding (1978) and
Kershaw and West (1991).

(4) Smooth margins. Kershaw and Riding (1978) identified an 'extended domical’ morphotype
with smooth margins created by non-enveloping laminae (Text-fig. 1 I). This can be modelled by

420

PALAEONTOLOGY, VOLUME 37

text-fig. 10. A chaetetid with sediment-induced ragged margins (from Kershaw and West, 1991, which
compare with forms modelled in Text-figures 4 and 5.

frequent, small sediment increments (e.g. forms in top left of Text-lig. 5), or by interference with
other individuals (Text-fig. 8).

(5) Bulbous versus pyramidal forms. Variation between generally pyramidal forms with broad
bases, and bulbous forms with relatively narrow bases, was documented by Kershaw and Riding
(1978). This variation can be modelled by varying the response to the sediment surface (compare
Text-fig. 6a, 7).

(6) Importance of hiatus. A comparison of Text-figures 4 and 5 demonstrates the importance of
an initial hiatus in sedimentation to permit growth. This is supported by the almost ubiquitous
observation of flat or concave bases of real stromatoporoids (see examples in Text-figs 9, 11): the
conical (convex-down) bases simulated in Text-figure 4 are rare or absent in stromatoporoids,
though common in corals. Furthermore, the prevalence of concave bases in stromatoporoids
suggests establishment on convexities (local highs) on the substrate, which would be sites of less
local sedimentation.

text-fig. II. An extended domical stromatoporoid,
with non-overlapping laminae (traced from photo-
graph of specimen from the Silurian of Gotland).
Compare with forms shown in Text-figures 5 and 9.
Scale bar represents 10 mm.

These six points of similarity allow the basic computer model to simulate most of the range of
observed morphologies described by Kershaw and Riding (1978; see, for example, their fig. 10). It
should prove possible to modify and control the computer model to simulate specific complex
fossils, and thus help improve understanding of an individual’s growth mode and history.

Discrepancies. The following points of discrepancy between simulations and real forms suggest that
we should retain some reservations about the total applicability of the model.

( 1 ) Some morphotypes are not realistically simulatable. The dendroid stromatoporoids do not
have the same geometry as those created by the model, as shown in Text-figure 7. The extended
domical morphotype with smooth margins and non-enveloping laminae has been simulated in two
ways (see point (4) above), but field evidence leaves some doubt as to whether real structures of this

SWAN AND KERSHAW: STROM ATOPOROID GROWTH

421

type can be attributed either to continuous sedimentation or to interference with adjacent
structures.

(2) Some simulations are unrealistic. It is a prediction of the basic model that many
stromatoporoids should have conical bases (Text-fig. 4), but these are rare in nature. As noted
above, the style of branching shown in Text-figure 7 is not represented in stromatoporoids, although
it is reminiscent of other organic forms. It is also possible that the modification of the model that
incorporates autonomous competition is not supportable as a stromatoporoid analogue.

IMPLICATIONS FOR THE INTERPRETATION OF STROMATOPOROIDS
Growth rate versus sedimentation rate

The results of the model confirm the interpretations of fossil morphologies as highly dependent on
sedimentation, particularly its rate and episodieity. The model also gives some idea of the relative
tolerance limits of stromatoporoids to sedimentation. An interesting and unexpected result in some
simulations (Text-fig. 5) was the manner in which growing stromatoporoids were initially able to
keep pace with sedimentation, but eventually become rather abruptly smothered, despite the
consistent pattern of sedimentation. A non-linear and apparently complex growth history can
therefore have a simple cause. However, sedimentation episodes will in reality have been variable
in frequency and amount, and stromatoporoid growth cannot be assumed to have been constant,
so survivorship of stromatoporoids under conditions of episodic sedimentation is likely to have
been haphazard.

Establishment of structures

We have observed that the model involving an initial hiatus (Text-fig. 5) produces more realistic
results than that involving no hiatus (Text-fig. 4). Indeed, the development of many typical
morphologies seems to be dependent on average sedimentation rate exceeding growth rate — a
terminal condition for which the initial hiatus is essential if the structures are to develop at all. There
are two possible interpretations of this.

( 1 ) Stromatoporoid growth may have been genuinely slow and unable to keep pace with typical
increments of sediment, so individuals would be immediately smothered unless they were ‘seeded’
on sites experiencing (for a period of time) near zero sedimentation. It is worth emphasizing that,
in this model, any stromatoporoid that successfully began growth while there was any sedimentation
would show signs of a conical base; the general rarity of these would specifically imply that the
initial hiatus was essential.

(2) The ‘hiatus’ may be only apparent and relative; there may be an initial rapid growth phase,
exceeding sedimentation rates, to allow the establishment of the structure. The lack of fossils having
the geometries simulated in Text-figure 4 would therefore be due to inadequacy of the basic model
of growth.

These alternatives are difficult to appraise; the observation of concave bases (convexities of
substrate) perhaps favours the former.

Level of skeletal integration

The comparability of simulations to fossil material suggests that stromatoporoids, and probably
also chaetetids, had some of the organizational attributes of the computer algorithm used here.
Specifically, it seems that each growth unit of a typical stromatoporoid, like each pixel in the model,
was largely autonomous in its susceptibility to local conditions; there is little reason to suppose that
directions, amounts and patterns of growth were under central control by the organism. However,
local autonomy was not absolute; there is no evidence of positive feedback between favourability
of position on the structure and growth rate, of the sort that produced the branching simulations
of Text-figure 7, so resources gained by favoured parts of the organism seem to have been

422

PALAEONTOLOGY, VOLUME 37

distributed. There may even have been a negative feedback mechanism to boost growth at incipient
recesses and hence maintain the smoothness of the surface.

The success of this model in producing growth forms which are analogous to real natural
structures suggests that it is a useful proxy for interpretation of highly integrated modular
organisms such as stromatoporoids. Similarities can also be observed between some of the
simulations and other organisms, particularly chaetetids, but also corals, stromatolites and
bryozoa. This paper has sought to describe and assess the computer model; forthcoming work will
explore its potential for improving understanding of specific specimens and palaeoenvironments.
Further experiments with computer models such as that presented here may lead to similar
interpretations of organizational level and growth strategies of organisms on the basis of the type
of computer algorithm and its parameter values.

Acknowledgements. We are grateful to Colin Stearn (McGill University) and an anonymous reviewer for
critical reviews of the manuscript.

REFERENCES

bosence, d. and waltham, d. 1990. Computer modeling the internal architecture of carbonate platforms.
Geology , 18, 26-30.

graus, r. R. and macintyre, i. g. 1976. Light control of growth form in colonial reef corals: computer
simulation. Science , 193, 895-897.

hartman, w. d. and goreau, t. f. 1970. Jamaican coralline sponges: their morphology, ecology and fossil
relatives. Zoological Society of London , Symposium , 25, 205-243.
hofmann, H. 1969. Attributes of stromatolites. Geological Survey Papers , Mines and Geology Branch , Canada ,
69-39, 1-58.

jackson, j. b. c. 1983. Biological determinants of present and past sessile animal distributions. 39-120. In
tevesz, m. and McCall, p. w. (eds). Biotic interactions in Recent and fossil benthic communities. Plenum Press,
New York, 837 pp.

kershaw, s. 1984. Patterns of stromatoporoid growth in level-bottom environments. Palaeontology, 27,
113-130.

1990. Stromatoporoid palaeobiology and taphonomy in a Silurian biostrome on Gotland, Sweden.
Palaeontology, 33, 681-705.

— and riding, r. 1978. Parameterization of stromatoporoid shape. Letliaia, 11, 233-242.

1980. Stromatoporoid morphotypes of the Middle Devonian Tor Bay Reef Complex at Long
Quarry Point, Devon. Proceedings of the Ussher Society, 5, 13-23.

— and west, r. r. 1991 Chaetetid growth form and its controlling factors. Letliaia, 24, 333-346.
labarbera, m. and boyajian, g. e. 1991. The function of astrorhizae in stromatoporoids: quantitative tests.

Paleobiology. 17, 121-132.

meyer, F. 1981. Stromatoporoid growth rhythms and rates. Science, 213, 894-895.

raup, d. 1966. Geometric analysis of shell coiling: general problems. Journal of Paleontology, 13, 1 178-1 190.
ross, J. r. p. 1987. New chaetetiform Trepostome bryozoa from the Upper Mississippian of the Western United
States. Palaeontology, 31, 551-566.

stearn, c. w. 1972. The relationship of the stromatoporoids to the sclerosponges. Letliaia, 5, 369-388.

— 1975. The stromatoporoid animal. Letliaia, 8, 89-100.

— 1984. Growth forms and macrostructural elements of the coralline sponges. Palaeontographica Americana,
54, 315-325.

1989. Intraspecific variability and species concepts in Palaeozoic stromatoporoids. Memoirs of the
Association of Australasian Palaeontologists, 8, 45-50.

vacelet, j. 1985. Coralline sponges and the evolution of the Porifera. Special Publications of the Systematics
Association, 28, 1-13.

wood, r. 1990. Reef-building sponges. American Scientist, 78, 131-156.

1991. Non-spicular biomineralisation in calcified demosponges. 322-340. In reitner, j. and keupp, h.
(eds). Fossil and Recent sponges. Springer-Verlag, Berlin, 595 pp.

— Zhuravlev, a. and debrenne, f. 1992. Functional biology and ecology of the Archaeocyatha. Palaios,
7, 131-156.

SWAN AND KERSHAW: STROM ATOPOROID GROWTH

423

young, G. a. and scrutton, c. t. 1991. Growth form of Silurian heliolitid corals: the influence of genetics and
environment. Paleobiology , 17, 369-387.

ANDREW R. H. SWAN

School of Geological Sciences
Kingston University
Kingston-upon-Thames KT1 2EE, UK

STEPHEN KERSHAW

Palaeobiology Research Unit
West London Institute

Typescript received 10 September 1993 Borough Road

Revised typescript received 18 January 1994 Isleworth TW7 5DU, UK

MORPHOLOGY OF ENCRUSTING AND FREE
LIVING ACE R VU LIN I D FO R AM IN I FER A:
ACER VULIN A, GYPSINA AND SOLENOMERIS

by CHRISTINE PERRIN

Abstract. The generic identification of acervulinids is especially difficult due to a confused systenratics.
However, this family is of major interest because it comprises the main encrusting reef Foraminifera which can
contribute significantly to the reef framework or build true reefs up to several kilometres in length. Their close
dependency on the substratum to which they are attached and their ability to develop various growth forms
result in an irregular morphology and arrangement of the chambers. This has certainly contributed to the
difficulty of defining accurate criteria for identification of genera and species. Moreover, the ability of the
Eocene acervulinid Solenomeris to build monospecific, kilometre-sized reefs has misled most previous workers
to consider it as a red alga. The geometrical characteristics of the test of the main acervulinid genera
( Acervulina , Borodinia , Gypsina , Solenomeris) are analysed and discussed, based on previous descriptions and
personal observations. This leads to some reliable and easily usable criteria for genera and species
identification. Solenomeris is very close to Acervulina but can be distinguished by the form of the juvenile.

The Acervulinidae includes sessile Foraminifera, often with an encrusting growth form, which are
able to contribute significantly to the reef framework or even to build true monospecific reef
biostrotnes (Plaziat 1984; Perrin 1987c/, 19876, 1987c, 1989, 1992; Plaziat and Perrin 1992) and
consequently is of major interest for palaeoenvironmental interpretation of Recent and Tertiary reef
facies. Like all attached organisms, acervulinids closely depend on the substratum they encrust; the
irregularities of the substratum surface influence the internal organization of the crust. Moreover,
like other reef-building organisms (e.g. scleractinians and Rhodophyceae), acervulinids have
developed various growth forms according to different environmental conditions (Perrin 1989,
1992).

The direct influence of the substratum on acervulinids and their tendency to develop different
growth forms are reflected in the irregularity of the geometry and arrangement of chambers. This
irregularity of the internal organization of the skeleton is probably one of the main causes of the
especially confused systematics of this family and has certainly contributed to the difficulty of
finding accurate criteria for the identification of genera and species. In particular, criteria for the
distinction between Acervulina and Gypsina have never been clearly defined. Moreover, the ability
of the Eocene Solenomeris to build monospecific, kilometre-sized reefs has led most previous
authors to consider them as red algae.

The most frequently quoted genera in reef environments are the fossil genus Solenomeris , and the
fossil to Recent genera Acervulina and Gypsina. This paper aims to provide reliable criteria for
generic and specific identification of acervulinids.

THE ACERVULINIDAE
Previous systematic studies of the family

The Acervulinidae was created by Schultze (1854) for Acervulina Schultze, 1854 (type species A.
inhaerens Schultze, 1854). This author also described two other species of the genus (A. glohosa and
A. acinosa) which do not seem to have been reported by subsequent authors, probably because of

| Palaeontology, Vol. .17, Part 2, 1994, pp. 425-458, 4 pls.|

© The Palaeontological Association

426

PALAEONTOLOGY, VOLUME 37

their very brief description and inadequate figures which show only the external appearance of the
test. Gypsina was described by Carter (1877) on the basis of a specimen described by Carpenter
(1876) and named Tinoporus vesicularis , itself a synonym of Orbit olina vesicularis. This last species
created by Parker and Jones (1860) therefore corresponds to the type species of Gypsina. Carter
(1877) also reported another species, Gypsina melobesioides, which he considered a synonym of
Polytrema planum previously created by himself (Carter 1876). This has resulted in much confusion
concerning both Acervulina and Gypsina. With the creation of Gypsina , Schultze’s work (1854)
concerning Acervulina became neglected and most of the subsequently described species were placed
in Gypsina and often in the Planorbulinidae. Brady (1884) referred Schultze’s species A. inhaerens
to Gypsina without explanation. The type species of Acervulina was first designated by Galloway
and Wissler (1927), and that of Gypsina by Cushman (1915). However, some authors have
considered Gypsina melobesioides and consequently Polytrema planum as the type species of Gypsina
(Loeblich and Tappan 1964; Moussavian 1989), though this species, sometimes named Gypsina
plana (Cushman et al. 1954), is considered by most authors as a synonym of Acervulina inhaerens
Schultze, 1854 (see Galloway and Wissler 1927) or its variety plana Hanzawa, 1931 (see Hanzawa
1931, 1957). Galloway and Wissler (1927) and Moussavian (1989) therefore regarded Acervulina
and Gypsina as synonyms since Polytrema planum , which they considered as the type species of
Gypsina , is an Acervulina.

Among the various subsequently described genera placed in this family some, like the unilocular
form Semseya Franzenau, 1893 (monotypic type species Semseya lameUata ), Pseudogypsina Trauth,
1918 (monotypic type species Pseudogypsina multiformis) and Borodinia Hanzawa, 1940 (monotypic
type species Borodinia septentrionalis ), remain extremely rare or very little known. Other genera
were created from existing species or varieties of Gypsina : Sphaerogypsina Galloway, 1933 (type
species Gypsina globulus Reuss, 1848), Discogypsina Silvestri, 1937 (type species Gypsina vesicularis
var. discus Goes, 1882), Planogypsina Bermudez, 1952 (type species Gypsina mastelensis Bursch,
1947). Ladoronia , created by Hanzawa (1957) as a subgenus of Acervulina (type species Acervulina
( Ladoronia ) vermicularis Hanzawa, 1957), was considered as a genus by Loeblich and Tappan
(1964).

Douville (1924) independently created the genus Solenomeris (type species Solenomeris O'Gormani
[s/c]) from an Eocene encrusting organism he identified as a red alga and which corresponds to the
Austrian Eocene form described by Trauth (1918) as Polytrema planum ( = Acervulina inhaerens var.
plana). Other species belonging to the same genus were described mainly as red algae (as
Solenoporacea or more rarely as coralline algae): S. douvillei Pfender, 1926, S. afonensis Maslov,
1956 and 5. pakistense Johnson and Konishi, 1960. However, several authors have placed
Solenomeris in the Acervulinidae (Hagn and Wellnhofer 1967; Hagn 1967, 1978, 1983; Moussavian
1984, 1989; Perrin 1987a, 19876, 1987c, 1992; Plaziat and Perrin 1992), some of them considering
it as a synonym of Gypsina (Hagn 1972, 1978, 1983; Moussavian 1984) or Acervulina (Moussavian
1989).

Identification of the Acervulinidae

The test of Acervulinidae consists of hyaline calcite and the walls are formed like those of the other
Rotaliina, by two calcified layers on both sides of an organic membrane (Hansen and Reiss 1971).
The test may be free or attached to a substratum (Schultze 1854; Galloway 1933; Loeblich and
Tappan 1964). Acervulinid growth is characterized by a spiral coiling of the early chambers,
followed by adult chambers arranged in one or several layers, without any apertures other than wall
pores, and without a canal system (Schultze 1854; Cushman 1950; Loeblich and Tappan 1964,
1984). However, there are some other characters common to the different genera of Acervulinidae,
especially the typical arrangement of the adult chambers alternating from one layer to the next in
multilayered tests.

Loeblich and Tappan (1964, 1984) suggested a stratigraphical range for this family from
Eocene to Recent. However, Cushman (1950) reported some acervulinids ( Acervulina and Gypsina)

PERRIN: ACERVULINID FO R A M I N I FE R A

427

text-fig. I Arrangement of chambers in
snbaxial section (left) and in transverse
section (right) in species of Acervulina.
A, A. inhaerens ; B, A. linearis., C,
A. (Ladoronia) vernacularism x 56.

m m

xz

W5&3E

:WEtr

table 1. Size of main features of the test of Acervulina inhaerens Schultze, 1854.

References and localities...

G. inhaerens
Yabe and
Hanzawa 1925
Ryukyu Id. Taiwan

var. plana
Hanzawa 1931
Japan

var. plana
Hanzawa 1
Micronesia

957

Test

Diameter (mm)

—

—

— —

— —

Thickness (mm)

—

—

— -

- -

Juvenile stages

Proloculus

Diameter (/mi)

—

—

— —

— —

Equatorial

Width (/mi)

—

—

- —

— —

chambers

Height (/mi)

—

—

— -

— —

Adult stage

Lateral

Width (/mi)

70-230

74-140

70-90 50-

230 56-100 56-130

chambers

Height (/mi)

—

32 — 43

5

18-47 23

Tangential wall

Thickness (/mi)

5

—

5 5

5 5

Lateral wall

Thickness (/mi)

5

—

5 5

5 5

Pores

Diameter (/mi)

5-7

—

5 5

5 5

Stolons

Diameter (/mi)

—

—

— —

428

PALAEONTOLOGY, VOLUME 37

from the Cretaceous. Moreover, specimens of Acervulina from the Upper Jurassic of Central Japan
have been described by Hanzawa (1939).

PRINCIPAL GENERA OF ACERVULINI DAE

Acervulina

Stratigraphical range. Acervulina is mainly reported from Cenozoic to Recent (Loeblich and Tappan 1964).
However, it is considered to have appeared before the end of the Mesozoic (Cushman 1950). Moreover,
Hanzawa (1939) described a variety of Acervulina inhaerens var. huzimotoi from an Upper Jurassic bioclastic
limestone in central Japan.

Type species. The type species Acervulina inhaerens (Text-fig. 1 ; Table 1), was described by Schultze (1854) from
a Recent shallow water specimen from the Ancona region of Italy. According to Schultze, it is characterized
by an attached or free-living test made up of a small number of chambers, each having a diameter of 60 pm.
The shell has a hyaline structure and is perforated by pores of 1 to 15 /an diameter. The material studied here
comes from Plio-Quaternary reefal limestone of Mururoa Atoll (French Polynesia).

Acervulina inhaerens

Juvenile stages (PI. 1, figs 1-2). The growth of juvenile forms can be divided into two main stages
(Perrin 1987 a): coiling development of the equatorial layer (or equatorial disc); and addition of
lateral chambers.

The equatorial layer (or equatorial disc) consists of spherical or subspherical chambers (average
diameter 80 //m) larger than the lateral chambers and formed in a planospiral arrangement around
the proloculus and the second peri-embryonar chamber. This equatorial disc is approximately
parallel to the substratum. The wall of the equatorial chambers appears imperforate but shows, like
the adult chambers, a dark median line separating two layers of fibrous hyaline calcite.

The lateral chambers present the same characteristics (shape and size) as the adult chambers.
They are arranged in successive layers around the equatorial disc, chambers alternating from one
layer to the next. In axial section, the three or four earlier successive layers of lateral chambers form
a slightly compressed oval (about 600 pm long and 200 pm high) showing a free bipolar growth.
Thus these earlier layers of lateral chambers intercalate between the equatorial layer and the
substratum. This indicates that their formation occurred before any attachment of the organism to
the substratum. Therefore, during this growth stage the organism was free-living and became
attached to a substratum only after the constitution of the third or fourth layer of lateral chambers.

EXPLANATION OF PLATE 1

Figs 1-8. Acervulina inhaerens Schultze, 1854. Mururoa atoll, Plio-Quaternary. 1, subaxial thin section showing
the ovoid of the juvenile stages; UPS Orsay F146; x85. 2, ovoid of the juvenile stages in subaxial thin
section; UPS Orsay F146; x 125. 3, arrangement of adult chambers in axial ultrathin section showing the
pores within the chamber roofs and some stolons in the lateral walls; UPS Orsay Ac4; x 160. 4, SEM of
axial section of adult chambers; the chamber roofs and floors are perforated and consist of two layers of
fibrous calcite developed on both sides of a median line; UPS Orsay Ac4'; x 490. 5, SEM showing the pores
in the chambers roofs and floors; the median line of the chamber wall is continuous through the pores; UPS
Orsay Ac7'; x 330. 6, SEM of tangential section through the chamber roofs showing the pores; UPS Orsay
Ac8'; x 175. 7, adult chambers in tangential ultrathin section showing the microstructure of the chamber
walls; the hyaline fibrous calcite of the chamber walls appears darker than the high-magnesian calcitic
cement filling the internal part of the chambers; UPS Orsay Acl ; x 160. 8, SEM of oblique section showing
the pores through the tangential walls; UPS Orsay Ac7'; x 270.

UPS, Universite Paris XI.

PLATE 1

PERRIN,

Acervulina

430

PALAEONTOLOGY, VOLUME 37

Adult stage. Contrary to the juvenile stages, the adult stage is characterized by a unipolar encrusting
growth. The juvenile settled on a rigid substratum and began to encrust it, adding successive layers
of chambers above and around the compressed ovoid. The thin crust followed the irregularities of
the substratum and, consequently, the arrangement of chambers is often irregular.

According to Hanzawa (1947, 1957), Acervulina inhaerens (Text-fig. 1 ; Table 1) is characterized
by upwardly arched roofs in axial section. However, in axial section, the adult chambers of the
Mururoa specimens have a flattened subhexagonal shape and are arranged one above the other in
successive layers with chambers alternating from one layer to the next (PI. 1, figs 3-4). Chamber
height is about 25-30 pm, while their average width is 70 /mi.

Each hexagonal chamber of a layer ‘n’ is delimited at its base by the roof of the underlying
chamber (layer ‘n-2’) and the lateral walls of the two chambers of the preceding layer (‘n-1’). Its
upper part is delimited by the roof and lateral walls, both newly formed, the lateral walls resting on
the lower lateral walls or, more rarely, on the roofs of the two chambers of the preceding layer
(PI. 1, fig. 5; Text-fig. 2).

text-fig. 2. Adult stage of Acervulina
showing the formation of vertical stacks
from the successive layers of chambers;
x 150.

In tangential section, the adult chambers have a rounded and irregular shape. Their diameter is
c. 60-80 /um. The shape and arrangement of the chambers is often more or less irregular due to
changes of growth direction during the development of the crust.

The chamber walls show the typical wall structure of lamellar hyaline Foraminifera (Hansen and
Reiss 1971; Haynes 1981; Loeblich and Tappan 1984): two layers of fibrous hyaline calcite are
developed on both sides of a dark median layer (PI. 1, fig. 4).

The flattened part of the wall forming the roof or floor of chambers or tangential wall is coarsely
perforated, the pores being 5-7 /mi in diameter (PI. 1, figs 6, 8). In ultrathin sections and in scanning
electron microscopy, the dark median layer of the tangential walls of Recent Acervulina appears to
be continuous through the pores (PI. 1, figs 4—5, 7).

The lateral walls show the same bilamellar structure bending downwards from the roof of the
chamber and leaning against the walls of the chambers of the underlying layer. However, the lateral
walls are imperforate.

The occurrence of stolons connecting adjacent chambers of the same layer has been reported by
Hanzawa (1957) but with some reservations, and later by Reiss and Hottinger (1984). Stolons
emerging from the lateral walls of adult chambers have been clearly observed in ultrathin sections
and in scanning electron micrographs of specimens from the cored wells of Muroroa Atoll (PI. 1,
fig. 3).

The variety plana is separated according to the larger size of its chambers (Yabe and Hanzawa
1925; Hanzawa 1931). On the other hand, the Jurassic variety huzimotoi (Hanzawa 1939) shows
smaller chambers and is also differentiated by its exceptionally thick sinuous walls (12-16 /im)
visible in tangential sections (Hanzawa 1939).

Other species

According to Hanzawa (1947, 1957), Acervulina is represented by two other encrusting species: Acervulina
( Acervulina ) linearis Hanzawa, 1947; and Acervulina (Ladoronia) vermicularis Hanzawa, 1957. Acervulina
linearis (Text-fig. 1 ; Table 2) was first described by Hanzawa (1947) from the Eocene of Micronesia and differs
from the type species by its non-arched roofs. The roofs of the adjacent flat chambers constitute a straight line
in axial section. The lateral walls are perpendicular or slightly oblique to the roof plane and do not show any
stolons (Hanzawa 1947, 1957). The subgenus Ladoronia created by Hanzawa (1957) is based on Acervulina

PERRIN: ACERVULINID FORAMINIFERA

43!

table 2. Size of main features of the test of Acervulina linearis Hanzawa, 1947.

References and localities...

A. linearis
Hanzawa 1947
New Britain

Marshall Id.

A. linearis
Hanzawa 1957
Micronesia

A. linearis
Hagn and
Wellnhofer 1967
Alps (Piaffing)

Test

Diameter (mm)

—

—

—

10 max.

Thickness (mm)

—

—

—

—

Juvenile stages

Proloculus

Diameter (/mi)

—

—

—

—

Equatorial

Width (/mi)

—

—

—

—

chambers

Height (/mi)

—

—

—

—

Adult stage

Lateral

Width (/an)

1 5-50

37-62

37-62 (21 1 max.)

—

chambers

Height (/an)

15-50

1 1-30

1 1-30

—

Tangential wall

Thickness (/mi)

5

5

5

4-5-9

Lateral wall

Thickness (/mi)

5

5

5

4-5-9

Pores

Diameter (/mi)

5-10

5-1 1

5-1 1

5

Stolons

Diameter (/mi)

—

—

—

—

( Ladoronia ) vermicularis from an Upper Oligocene-Aquitanian limestone of Micronesia (Text-fig. 1 ; Table 3).
This species possesses chambers larger than Acervulina inhaerens and characterized by their elongated sinuous
shape in tangential section. These chambers communicate by way of stolons within the same layer (Hanzawa
1957). The juvenile stages of this species have been described by Hanzawa ( 1957, pp. 68-69) as a ‘raspberry-
like embryonic apparatus’ encircled by ‘two or three annuli of arcuate chambers’. The nepionic chambers
show thick roofs (> 100 //m) within which vertical nontubulous pillars are embedded. The neanic chambers
are large and vermicular-shaped and have thick vertical lateral walls pierced by large stolons (Hanzawa 1957).

Generic characteristics of Acervulina

The juvenile stages of Acervulina are free and characterized by the formation of a three-layered
ovoid. The equatorial layer consists of a planospiral arrangement of subspherical large chambers
around the proloculus and the second periembryonar chamber. This first stage is followed by the
addition of layers of lateral chambers on each side of the equatorial disc, forming the ventral and
dorsal zones.

The adult stage is attached to a substratum by the ventral face and is characterized by a unipolar
growth of the dorsal zone. The adult chambers are subhexagonal in axial section and show rounded
shapes in tangential section. They are arranged in successive layers with chambers alternating from
one layer to the next one. Tangential and lateral walls show a thickness of a few microns (5-7 pm).
Chambers from successive layers communicate by way of perforations of the roofs (tangential walls)
of the chambers.

The distinction of the three species of Acervulina appears to be mainly based on the size and the
shape of chambers (Text-fig. 1).

Growth pattern

The genus Acervulina includes encrusting forms with very different thicknesses: from less than one
millimetre to several centimetres. Acervulina ( Ladoronia ) vermicularis generally forms a thin crust
(Hanzawa 1957), whereas Acervulina inhaerens constitutes millimetre-thick as well as decimetre-
thick crusts (Yabe and Hanzawa 1925; Galloway and Wissler 1927; Hanzawa 1939; Hottinger
1983; Reiss and Hottinger 1984; Perrin 1987 b, 1989, 1990, 1992). Moreover, A. inhaerens can

432

PALAEONTOLOGY, VOLUME 37

table 3. Size of main features of the test of Acervulina ( Ladoronia ) vermicularis Hanzawa, 1957.

Reference and locality...

A. ( Ladoronia ) vermicularis
Hanzawa 1957
Micronesia

Test

Diameter (mm)
Thickness (mm)

Juvenile stages

Proloculus Diameter (//m) 68

Equatorial Width (/mi)

chambers Height (/mi)

Adult stage
Lateral
chambers
Tangential wall
Lateral wall
Pores
Stolons

Thickness (/mi) 9

Diameter (/mi) 5

Diameter (/mi) 14

Thickness (/mi) 9-24

Width (//m)
Height (/mi)

develop different morphologies according to environmental conditions (Hottinger 1983; Reiss and
Hottinger 1984; Perrin 1989, 1992), the most frequently observed being crusts a few millimetres
thick, extending over a surface area of a few square centimetres. Nodules or macroids (sensu
Hottinger 1983) with a centimetre to decimetre diameter have been described at the base of reef-
slopes (Chapman 1900; Logan et al. 1969; Hottinger 1983; Reiss and Hottinger 1984; Dullo et al.
1990). Moreover, in some cored wells of the Mururoa Atoll, Acervulina inhaerens is responsible for
boundstones formed by decimetre-thick crusts which characterize the palaeoenvironments of deeper
external reef-slopes (Perrin 1989, 1990, 1992).

Borodinia

Stratigraphical range. This genus was created by Hanzawa (1940) for encrusting Foraminifera from the
Aquitanian limestone of Kita Daito Jima (China Sea) and was later recorded by the same author from the
Aquitanian limestones of Micronesia (Hanzawa 1957). Being very rarely reported, its stratigraphical range is
especially difficult to establish.

Type species. According to Hanzawa's description of the type species, B. septentrionalis Hanzawa. 1940, from
the Aquitanian limestone of Kita Daito Jima drill cores, this encrusting foraminifer forms a layer of one or
more zones of chambers. The outer wall of the shell is 37-75 /mi thick and shows coarse pores of 11 /mi
diameter. The chambers are irregularly arranged and each of them consists of an arched tangential wall and
lateral walls of the same 12-25 /mi thickness. Both lateral walls of each chamber are pierced by large stolons
of 37 //m diameter.

Borodinia septentrionalis

Juvenile stages. The organization of the earlier chambers is similar to Planorbulina and shows a
planospiral arrangement (Hanzawa 1957).

Adult stage. The adult chambers are very irregularly arranged in alternating successive layers
(Hanzawa 1957; Loeblich and Tappan 1964). In transverse sections, some chambers are typically
spatuliform and the adjacent chambers communicate with each other by several stolons (Text-fig. 3;
Table 4). The roofs of chambers are especially thick (140 //m) in comparison with the average size

PERRIN: ACER VU LIN I D FORAMINIFERA

433

text-fig. 3. Arrangement of chambers in subaxial section (left), showing the finely cribate roofs, and in
transverse section (right), with numerous stolons, in Borodinia (after Hanzawa 1957); x 70.

table 4. Size of main features of the test of Borodinia septentrionalis Hanzawa, 1940 (after Hanzawa 1940,
1957).

Reference and locality...

Bordinia septentrionalis
Hanzawa 1957
Micronesia

Test

Diameter (mm)

—

Thickness (mm)

—

Juvenile stages

Proloculus

Diameter (/nn)

—

Equatorial

Width (//nr)

—

chambers

Height (/nn)

—

Adult stage

Lateral

Width (/nn)

120-210

chambers

Height (/nn)

122

Tangential wall

Thickness (/nn)

140

Lateral wall

Thickness (/nn)

20

Pores

Diameter (/nn)

9

Stolons

Diameter (/nn)

14

of chambers ( 100-200 //m x 122 /mr), while the thickness of lateral walls is 20 /nn (Table 4). The roof
wall is finely perforated (Hanzawa 1957).

Generic characteristics of Borodinia

Borodinia is characterized by a planospiral arrangement in the juvenile stage. The adult chambers
show a very irregular shape, often spatuliform in tangential section, and they communicate with
adjacent chambers by means of stolons. Tangential walls of adult chambers have a thickness of
more than 100 //m and are perforated by pore canals (Hanzawa 1940, 1957; Text-fig. 3).

434

PALAEONTOLOGY, VOLUME 37

text-fig. 4. Different species of Gypsina in vertical section, a, G. vesicularis ; b, G. globulus ; c, G. saipanensis;
D, G. marianensis; E, G. squamiformis; F, G. mastelensis ; x 26.

table 5. Size of main features of the test of Gypsina vesicularis (Parker and Jones, 1860).

References and localities...

G. vesicularis
Bursch 1947
Mollusk Id

G. vesicularis
Hanzawa 1957
Micronesia

Test

Diameter (mm)

2-9 max.

1-3-1 -7

Thickness (mm)

0-65

0-56-0-79

Juvenile stages

Proloculus

Diameter (pm)

—

47-78

Equatorial

Width (//m)

—

44-70

chambers

Height (/mi)

—

44-140

Adult stage

Lateral

Width (/mi)

80

60-82

chambers

Height (/mi)

5-20

5-25

Tangential wall

Thickness (/mi)

5-10

10-14

Lateral wall

Thickness (/mi)

10-15

—

Pores

Diameter (/mi)

5

5

Stolons

Diameter (/mi)

—

—

Growth pattern

The specimens described by Hanzawa (1940, 1957) are encrusting forms but no indication about the
extent and the thickness of the crust is given.

Gypsina

Stratigraphical range. This genus is generally reported from the Eocene to Recent (Bursch 1947; Hanzawa
1957; Loeblich and Tappan 1964), but Cushman (1950) considered it as probably dating back to the
Cretaceous.

Type species. Gypsina was created by Carter (1877) for Tinoporus vesicularis Carpenter, 1876, a synonym of
Orbitolina vesicularis Parker and Jones, I860. The type species of Gypsina is therefore Gypsina vesicularis
(Parker and Jones, 1860), from Recent coral-reef sediment of Australia.

PERRIN: ACERVULINID FORAMIN1FERA

435

The type-specimen described by Parker and Jones (1860) has a slightly conical test of 2-5 mm diameter
consisting of vesicular chambers (tangential section), some of them showing polygonal shapes, arranged
alternately concentrically (axial section). The equatorial disk, referred to as the ‘primary disk’, is generally
covered by additional layers of chambers (dorsal lateral chambers) and the ‘umbilicus is filled up so that the
base of the cone is almost flat’ (ventral lateral chambers). The walls have coarse pores referred as
‘pseudopodial passages' (Parker and Jones 1860, pp. 31 32).

Gypsina vesicularis

This species (Text-fig. 4a; Table 5) has a discoidal test with a convex or conical shape and a
thickened round periphery. The proloculus is spherical and located at the centre of the equatorial
layer (Bursch 1947; Hanzawa 1957). The equatorial chambers have arched shapes and larger size
than lateral chambers. They show stolons at the edges and perforations in the arched parts of their
walls (Bursch 1947). The lateral chambers surround the equatorial disc and are symmetrically
arranged on both sides of the median (equatorial) layer (Brady 1884; Hanzawa 1931, 1957; Bursch
1947; Cushman et al. 1954). They show the same shape and the same arrangement as the lateral
chambers of Acervulina. Each chamber communicates with the adjacent chambers through stolons
at the base of lateral walls, and with the chambers situated above and below through perforations
in the tangential wall (Bursch 1947; Hanzawa 1957).

Gypsina discus Goes (see Bursch 1947) is now considered as a variety of the type species (Hanzawa
1957) characterized by its discoidal shape. It is the type species of Discogypsina Silvestri, 1937.

Other species

According to Hanzawa (1931, 1957) and Bursch (1947), Gypsina is represented by at least five other species:
G. globulus (Reuss, 1848); G. squamiformis Chapman, 1900; G. mastelensis Bursch, 1947; G. saipanensis
Hanzawa, 1957; G. marianensis Hanzawa, 1957.

Gypsina globulus (Text-fig. 4b; Table 6), which was first described by Reuss (1848) as Ceriopora globulus , is
characterized by its spherical to subovoidal shape and more regular arrangement of the chambers, smaller
chambers and finer pores than other Gypsina species. It also lacks the median layer (Brady 1884; Hanzawa
1931 ; Bursch 1947; Cushman et al. 1954). Bursch (1947) described a megalospheric juvenarium composed of
two embryonar chambers and two auxiliary chambers. Stolons occur within the lateral walls of adult chambers

table 6. Size of main features of the test of Gypsina globulus (Reuss, 1848).

References and localities...

G. globulus
Hanzawa 1931
Japan

G. globulus
Bursch 1947
Mollusk Id.

G. globulus
Hanzawa 1957
Micronesia

Test

Diameter (mm)
Thickness (mm)

2-25

0-5-1

1 -4-2-0

Juvenile stages

Proloculus

Diameter (pm)

—

37

Equatorial

Width (//m)

—

—

chambers

Height (/mi)

—

— -

Adult stage

Lateral

Width (pm)

—

60

150-180

chambers

Height (/mi)

—

50

25-50

Tangential wall

Thickness (//m)

—

3-6

10-15

Lateral wall

Thickness (/mi)

—

5-8

—

Pores

Diameter (/mri)

—

5

5

Stolons

Diameter (/mi)

—

5

—

436

PALAEONTOLOGY, VOLUME 37

table 7. Size of main features of the test of Gypsina squamiformis Chapman, 1900.

Reference and locality...

G. squamiformis
Bursch 1947
Mollusk Id

Test

Diameter (mm) 1

Thickness (mm)

Juvenile stages

Proloculus
Equatorial
chambers
Adult stage
Lateral
chambers
Tangential wall
Lateral wall
Pores
Stolons

Diameter (/un)
Width (//m)
Height (/mr)

Width (//nr)
Height (//m)
Thickness (/mr)
Thickness (//m)
Diameter (//m)
Diameter (/mr)

120

4-5

30 max.

table 8. Size of main features of the test of Gypsina mastelensis Bursch, 1947.

Reference and locality...

G. mastelensis
Bursch 1947
Mollusk Id

Test

Diameter (mm)

1-2 (2-2 max.)

Thickness (mm)

OT-O-2 (0 3 max.)

Juvenile stages

Proloculus

Diameter (/nn)

—

Equatorial

Width (/mr)

—

chambers

Height (/mr)

40-70

Adult stage

Lateral

Width (/mr)

90 max.

chambers

Height (/mr)

—

Tangential wall

Thickness (//nr)

5-10

Lateral wall

Thickness (//m)

5-10

Pores

Diameter (/mr)

5

Stolons

Diameter (//nr)

10-15 (20 max.)

(Bursch 1947; Reiss and Hottinger 1984). Specimens from cored wells from Mururoa Atoll show similar
chamber shapes as Acervulina. Chamber size is 60 /mr. Reiss and Hottinger (1984) noted a more regular shape
and arrangement of the chambers in G. globulus than in Acervulina inhaerens. Sphaerogypsina Galloway, 1933,
was created for the species Gypsina globulus (Reuss, 1848).

Gypsina squamiformis (Text-fig. 4e; Table 7) is easily separated from the other Gypsina species since it is only
formed by a single layer of arched chambers (Chapman 1900), which increase in size at the periphery of the
encrusting test (Bursch 1947). Recent specimens described by Chapman (1900) from Funafuti have a test
diameter of about 3-4 millimetres. The tangential walls of these chambers are perforated and large stolons
allow communication between adjacent chambers (Bursch 1947). This species has been considered as a distinct
genus, Planogypsina Bermudez, 1952.

PERRIN: ACER V U L I N I D FORAMINIFERA

437

table 9. Size of main features of the test of Gypsina saipanensis Hanzawa, 1957.

Reference and locality...

G. saipanensis
Hanzawa 1957
Micronesia

Test

Diameter (mm)

1 -7-2-6

Thickness (mm)

0-25-0-60

Juvenile stages

Proloculus

Diameter (/mi)

40

Equatorial

Width (/mi)

—

chambers

Height (/mi)

98-294

Adult stage

Lateral

Width (/mi)

50

chambers

Height (/mi)

50

Tangential wall

Thickness (/mi)

—

Lateral wall

Thickness (/mi)

—

Pores

Diameter (/mi)

—

Stolons

Diameter (/mi)

—

table 10. Size of main features of the test of Gypsina marianensis Hanzawa, 1957.

Reference and locality...

G. marianensis
Hanzawa 1957
Micronesia

Test

Diameter (mm)

2-7

Thickness (mm)

0-6

Juvenile stages

Proloculus

Diameter (/mi)

120

Equatorial

Width (/mi)

80-94

chambers

Height (/mi)

120

Adult stage

Lateral

Width (/mi)

57

chambers

Height (/mi)

57 (vent.) 10 (dors.)

Tangential wall

Thickness (/mi)

—

Lateral wall

Thickness (/mi)

—

Pores

Diameter (/mi)

10 (dors.)

Stolons

Diameter (/mi)

14 (vent.)

G. mastelensis (Text-fig. 4f; Table 8), described by Bursch ( 1947). is composed of a basal layer of equatorial
chambers, adherent to the substratum, and a dorsal arched zone. The juvenarium, situated at the centre of the
equatorial layer, comprises three equatorial chambers. The next equatorial chambers are radially arranged
around the embryonar chambers and their height generally increases towards the periphery of the test. There
are also communications between the equatorial chambers through roof pores and stolons. The lateral
chambers communicate with each other and with the equatorial chambers through stolons. Bermudez (1952)
created the new genus Hemigypsina for this species.

The test of G. saipanensis (Text-fig. 4c; Table 9) may present a concavo-convex, or plano-concave, or
discoidal shape, which gives it a uniform thickness or a thickened periphery. The test comprises an equatorial
layer, two or three smaller chambers irregularly arranged on the ventral side, and a single layer of chambers

438

PALAEONTOLOGY, VOLUME 37

table 11. Age, localities and systematic assignments of Solenomeris by previous authors.

Systematic

Reference

Taxon

Age

Growth form

Locality

position

Trauth 1918

P. planum

Eoc.

Nodules

Austria

Acervulinidae

Douville 1924

S. ogormani

1. Eoc.

Nodules

S. France

Lithothamniae

Douville and

S. ogormani

1. Eoc.

Nodules

S. France

Calcareous algae

O'Gorman 1924

Pfender 1926

S. douvillei
S. sp.

l. Eoc.

m. Eoc.

Encrusting

N. Spain
Italy

Hydrozoan

Rao and
Varma 1953

S? douvillei
non Pfender

1. Eoc.

Encrusting

Pakistan

Melobesiae

Maslov 1956

S. afonensis

1. Eoc.

7

Abkhazie

Similar to disco-
cyclines and
stromatopores

Elliott 1960

S. douvillei
S. ogormani

Palaeoc.
1. Eoc.

7

Iraq

Solenoporaceae

Johnson and

S. pakistense

1. Eoc.

?

Pakistan

Solenoporaceae

Konishi 1960

Massieux 1961

S. douvillei

1. Eoc.

Reef

S. France

lncertae sedis
algae

Schalekova 1963

S. douvillei

m. Eoc.

7

Slovakia

Calcareous algae

Elliott 1964

S. ogormani
S. douvillei

Palaeoc. to
1. Eoc.

7

Iraq

Solenoporaceae

De Zanche 1965

Solenomeris

Eoc.

Fragments

N. Italy

Melobesiae

Toumarkine

Solenomeris

1. Eoc.

Reefs

S. France

—

1966; 1967

Hagn 1967

Solenomeris

1. Eoc.

7

Alps

Foraminifera

Hagn and

Solenomeris

u. Eoc.

Encrusting

Alps

Acervulinidae

Wellnhofer 1967

Boulanger and

Solenomeris

m. Eoc.

Encrusting

S. France

Algae

Poignant 1969

Terry and

—

—

7

Libya

Solenoporaceae

Williams 1969

Poignant and
Du Chaffaut 1970

S. ogormani

Palaeoc.

7

France

lncertae sedis
algae

Hagn 1972

G. ogormani

Palaeoc. -Eoc.

Encrusting

Alps

Gypsina

Samuel et al. 1972

S. sp.

Maas. -Palaeoc.

7

Carpathians

Algae

Massieux 1973

S. douvillei

1. Eoc.

Reef and crusts

S. France

lncertae sedis
algae

Poignant 1974

Solenomeris

u. Cret.
1. Oligo.

7

7

Algae

Poignant and

S. ogormani

u. Mio.

—

S. France

Solenoporaceae

Blanc 1974

Tambareau and

Solenomeris

1. Eoc.

Encrusting

S. France

—

Villatte 1974

Poignant 1976

Solenomeris

Eoc.

7

Spain

Red algae

Deloffre et al. 1977

S. sp.

Senonian

Fragments

Iran

Red algae

De Zanche et ai

Solenomeris

Eoc.

Fragments

N. Italy

Solenoporaceae

1977

Orszag et al. 1977

S. douvillei

Mio.

Encrusting

T urkey

—

Poignant 1977

S. sp.

Palaeoc.

Fragments

Paris B.

Red algae

Wray 1977

Solenomeris

Palaeoc.

Encrusting

Libya

Solenoporaceae

Gaemers 1978

Lithothamnium

1. Eoc.

' ridges ’

N. Spain

Melobesieae

Hagn 1978

G. ogormani

Palaeoc.-l. Eoc.

Encrusting

Alps

Gypsina

PERRIN: ACERVULINID FORAMINIFERA

439

TABLE 1 1 . (C0nt.)

Reference

Taxon

Age

Growth form

Locality

Systematic

position

Dieni et al. 1979

S. ogormani

Palaeoc.-l. Eoc.

Fragments

Sardinia

Rhodophyceae

Hagn and

G. ogormani

Eoc.

Encrusting

Alps

Gypsina

Moussavian 1980

Tappan 1980

Soienomeris

—

—

—

Solenoporaceae

Beckmann et al.

S. sp.

Palaeoc.

Fragments

Italy

Solenoporaceae

1982

Gravello and

Soienomeris

Eoc.

Fragments

Italy

Calcareous algae

Ungaro 1982

Hagn 1983

G. ogormani

1. and m. Eoc.

Encrusting

Austria

Gypsina

Moussavian 1984

G. ogormani

u. Oligo.

—

Alps

Gypsina

Plaziat 1984

S. douviiiei

1. Eoc.

Reef

S. France
N. Spain

Probably encrusting
Foraminifera

Perrin 1985

Soienomeris

1. Eoc.

Reef

S. France
N. Spain

Acervulinidae

Betzler 1987

Soienomeris

1. Eoc.

?

N. Spain

Red algae

Brugnatti and

Soienomeris

Eoc.

Fragments

Italy

Calcareous algae

Ungaro 1987

Eichenseer 1987

Soienomeris

1. Eoc.

‘ Ridges 7

N. Spain

Red algae

Perrin 1987a, b , c
Perrin and

Soienomeris

1. Eoc.

Reef

S. France
N. Spain

Acervulinidae

Plaziat 1987

Eichenseer 1988

Soienomeris

1. Eoc.

Crusts macroids
' ridges 7

N. Spain

?

Betzler 1989

Soienomeris

1. and m. Eoc.

Encrusting

N. Spain

Red algae

Moussavian 1989

Soienomeris

Eoc.

—

Alps

Acervulina ogormani

Perrin 1992
Plaziat and

Soienomeris

1. Eoc.

Crusts, macroids,
and reefs

S. France
N. Spain

Acervulinidae

Perrin 1992

on the dorsal side. The equatorial layer includes a spherical proloculus surrounded by chambers with their
diameter increasing towards the periphery (Hanzawa 1957).

Gypsina marianensis (Text-fig. 4d; Table 10), proposed by Hanzawa (1957), also has a plano-convex or
concavo-convex shaped test which consists of three clearly distinct zones: median, dorsal and ventral. The
median arched zone includes a spherical proloculus which is localized at the apex and a single layer of chambers
with a spiral arrangement in the nepionic stage and an annular arrangement in the later neanic stage. Within
this zone, each chamber shows two stolons at its opposite extremities. The dorsal zone is made of two or three
layers of quadrangular depressed chambers (in axial section). The ventral zone is formed by several layers of
chambers which tend to fill the hollow of the cone made by the median zone. The ventral chambers also show
a depressed shape and generally a larger size than the median and dorsal chambers. The number of dorsal
layers is always less than ventral layers (Hanzawa 1957).

Generic characteristics of Gypsina

This genus is represented by encrusting or free-living species. The proloculus has a spherical shape.
The type species Gypsina vesicularis has a three-zoned test but in other species some zones may be
absent. The dorsal zone, when present, does not show a well-developed unipolar growth but is
represented by only a few layers of chambers. The chambers have walls a few microns in thickness
and a perforated tangential wall. Species are discriminated according to the general shape of the test
and the number and geometry of the different layers of chambers.

440

PALAEONTOLOGY, VOLUME 37

Growth pattern

Species of Gypsina may have free, attached or encrusting habits. G. globulus is a typical free-living
species probably living on plants (Reiss and Hottinger 1984), while G. squamiformis develops thin
crusts on hard substrates (Chapman 1900; Yabe and Hanzawa 1925; Bursch 1947; Reiss and
Hottinger 1984). Several species show diversity in their growth habits. For example, G. marianensis
shows a conical plano-convex or concavo-convex shaped test which is attached to the substratum
but may sometimes extend across that surface and become encrusting (Hanzawa 1957). Bursch
( 1947) described specimens of G. mastelensis , a normally sessile and encrusting species, which adopts
an irregular growth morphology when it becomes detached from its substratum during development.

Solenomeris

Identification of Solenomeris as a Foraminifera. Solenomeris has been referred to diverse groups of
organisms, especially Rhodophyceae (Table 1 1). However, Trauth (1918) identified nodular forms
of this organism from the Eocene of Austria (Province of Pongau) as the encrusting Foraminifera
Polytrema planum ( = Acervulina inhaerens var. plana according to Hanzawa 1957). Nevertheless,
Trauth’s publication was overlooked by subsequent authors until Hagn (1967). Thus, Douville
(1924) created the genus Solenomeris , with the type species Solenomeris O'gormani [v/cj from the
Early Eocene of the Bearn region (SW France). This author linked S. ogormani to an isolated
branch of Lithothamniae (Douville 1924, p. 170) while emphasizing the greater size of Solenomeris
‘cells' compared with Lithothamniae. He did not assign the new genus to Solenoporaceae because
of the difference in ‘cell’ shape in transverse section. In hand-specimens, these Solenomeris form
‘small calcareous balls more or less mammilated with a smooth surface’ (p. 169).

From the vicinity of Pau, Douville and O'Gorman (1924) described Eocene ‘reefs' built up by
Solenomeris associated with Alveolina , polystomelles and corals. However, these calcareous balls
seem to be non-reefal material resedimented in deeper parts of the basin (Plaziat and Perrin 1992).

Pfender (1926) described another Solenomeris , S. douvillei , from the Early Eocene of the Spanish
Pyrenees (Camarasa, Province of Lerida). According to Pfender, this species differs from the type
species in its encrusting growth alternating with red algal crusts and never constitutes isolated
calcareous nodules. Pfender (1926) recorded also Solenomeris sp. which forms ‘globular masses’ in
the Middle Eocene of Vicentin (northern Italy). She compared Solenomeris with hydrozoans owing
to the similar aspect of the ‘tissue’.

Rao and Varma (1953) reported a new species, 5? douvillei (non Pfender), from the Early Eocene
of Pakistan. According to these authors, this species has an encrusting growth habit differing from
S', ogormani with smaller sized ‘cells’. The place of Solenomeris among melobesians is paradoxically
based on ‘cell’ sizes and on the more characteristic presence of conceptacles including reproductive
cells (Rao and Varma 1953, p. 22); these structures correspond in fact to juvenile stages. The species
of Rao and Varma (1953) was later named Solenomeris pakistense by Johnson and Konishi (1960).

Maslov (1956) described the new species Solenomeris afonensis (Early Eocene of Abkhazie) and
noted the similarity of Solenomeris ‘cells’ with those of discocyclines and stromatopores.

The two species Solenomeris ogormani and S. douvillei were recorded in the Paleocene and Early
Eocene of Iraq by Elliott (1960) who placed them among Solenoporaceae. He argued that the algal
nature of Solenomeris is supported by the existence of unipore conceptacles similar to those of
Lithothamniae (Elliott 1964). However, Elliott (1964) underlined the rarity of reproductive organs
and considered the conceptacles described by previous authors (Rao and Varma 1953; Maslov
1956) as Foraminifera belonging to the genus Bullopora subsequently encrusted by Solenomeris.

Most Solenomeris then have been misidentified as calcareous algae (Schalekova 1963; Boulanger
and Poignant 1969; Poignant and Du Chaffaut 1970) and classified among Rhodophyceae (Samuel
etal. 1972; Poignant 1975, 1977; Dieni et at. 1979; Deloffre etal. 1977; Garavello and Ungaro 1982)
or more precisely among Solenoporaceae ( De Zanche 1 965 ; Terry and Williams 1 969 ; Poignant and
Blanc 1974; De Zanche et al. 1977; Wray 1977; Tappan 1980) or among Corallinaceae (Gaemers
1978).

PERRIN: ACERVULINID FORAMINIFERA

441

Although some authors noted the resemblance of Solenomeris to discocyclines (Maslov 1956) or
orbitoids (De Zanche 1965), Hagn (1967) was the first, after Trauth (1918), to refer Solenomeris
from the Early Eocene (‘Ilerdian’) of the Bavarian Alps to the Foraminifera, comparing it with the
acervulinid genera Acervulina and Gypsina ( Hagn and Wellnhofer 1967). The genus Solenomeris was
placed in the Family Acervulinidae and considered as a possible synonym of Gypsina (Hagn and
Wellnhofer 1967).

Hagn subsequently used this new combination for Douville’s species Gypsina ogormani (Hagn
1972, 1978, 1983; Hagn and Moussavian 1980; Moussavian 1984) because of the presence of a
juvenarium similar to that of Gypsina (Hagn 1972, pp. 1 16-1 17), but without giving any illustration.
Specimens from the Palaeocene and Eocene of the Bavarian and Austrian Alps show only
encrusting growth forms, reworked in detrital facies (pebbles and blocks). Moussavian (1989)
attributed Solenomeris to the genus Acervulina since he considered Gypsina to be a synonym of
Acervulina. However, Brugnatti and Ungaro (1987) and Barbin et al. (1989) identified Solenomeris
from the Middle Eocene of Northern Italy as Gypsina.

In spite of Hagn’s publications, most other authors have not considered Solenomeris as a
foraminifer and it is persistently placed among the red algae. So, in Tappan's synthesis of plant
protists (1980), Solenomeris appears in the Family Solenoporaceae (p. 140).

Solenomeris douvillei from the Early Ypresian (‘Ilerdian’) of the Corbieres (Southern France)
forms reef-sized build-ups. It has been placed among incertae sedis algae by Massieux ( 1961 , 1973),
while the reefal Solenomeris from the Southern Pyrenees (Spain) were misidentified by Gaemers
(1978) as Lithothamnium ridges. Beckmann et al. (1982) recorded Solenomeris from the Palaeocene
of the Bergamo region (northern Italy) as Solenoporaceae but pointed out the problem of the algal
nature of Solenomeris by referring to Hagn and Wellnhofer (1967).

The taxonomic problem of Solenomeris was pointed out by Plaziat (1984) without any conclusion
in his palaeogeographical study of the Pyrenean region in which the different Eocene reefs of the
Corbieres and North-East Spain were described and interpreted for the first time from a
palaeoecological perspective (Plaziat 1984).

Comparison of the structural features of the Solenomeris test with those of Acervulina from Plio-
Quaternary reefs of Mururoa Atoll (French Polynesia) led to the conclusion that Solenomeris
belonged to the Family Acervulinidae (Perrin 1985, 1987a). This was based on the identification of
the juvenile stages with a geometry and arrangement of chambers which confirms assignation to the
Family Acervulinidae (Perrin 1987a, 19876; Perrin and Plaziat 1987). However, these juvenile stages
were previously described and illustrated but misinterpreted as conceptacles (Rao and Varma 1953;
Boulanger and Poignant 1969; Poignant and Du Chaffaut 1970; Poignant and Blanc 1974) or as the
Foraminifera Bullopora sp. (Elliott 1964) or incertae sedis (Massieux 1973, pi. 26, fig. 5).

Stratigraphical range. Solenomeris is mainly reported from the Eocene, more rarely from the Palaeocene
(Elliott 1960, 1964; Hagn 1972, 1978; Samuel et al. 1972; Beckmann et al. 1982). References to Late
Cretaceous Solenomeris are rare. In the Senonian of Central Iran, Deloffre et al. (1977) reported fragments of
Solenomeris , which however seem to be reworked. On the other hand, S. ogormani from the Upper
Maastrichtian of Roquefort-Creon (Central Aquitaine) shows, according to Poignant and Blanc (1974), a
significant build-up role. The Solenomeris sp. reported by Samuel et al. (1972) from the Maastrichtian and
Montian-Thanetian of the Carpathians seem to grow in an encrusting habit. Poignant (1975) gave the
stratigraphical range of Solenomeris as Upper Cretaceous to Upper Oligocene, with an acme during the
Eocene. The youngest specimens of Solenomeris are from the Turkish Miocene (Poisson and Poignant 1975;
Orszag-Sperber et al. 1977).

The development of Solenomeris reefs is so far known only from the Early Eocene (Lower Ypresian) of the
Pyrenean region (Plaziat 1984; Plaziat and Perrin 1992).

Type species. The name of the type species Solenomeris O'Gormani described by Douville (1924) has been
amended to S. ogormani. According to Douville (1924), Solenomeris ogormani is characterized by its growth
as independent nodules and by a ‘cell’ ranging from 35 to 50 pm. The internal structure of the test consists of
successive concentric layers of ‘cells’, some sections showing radial series of ‘cells’. In transverse sections, these

442

PALAEONTOLOGY, VOLUME 37

have an irregularly hexagonal shape and thin walls, while in axial sections, they are arranged in vertical series,
alternating from one series to the adjacent one, this resulting in hexagonal ‘cells’. Some ‘cells' are flattened
(axial sections) and have been interpreted as resulting from slowed growth.

Solenomeris ogormani

The detailed reconstruction of Solenomeris development and its assignation to Foraminifera result
from complementary observations using optical microscopy (thin and ultrathin sections) and
scanning electron microscopy (fresh fractures or polished surfaces briefly etched with formic acid)
of Ypresian specimens from the Pyrenees ( Solenomeris ogormani) and from Plio-Quaternary reefs
of the Mururoa atoll {Acervulina inhaerens , identification by M. Neumann in Repellin 1977; Perrin
1985, 1987a).

text-fig. 5. Juvenile stages of Solen-
omeris and beginning of encrusting adult
stage (after Perrin 1987a). a, equatorial
layer; B, ovoid stage in axial section (left)
and in transverse section (right); c, test
attached to a substratum and devel-
opment of the crust; x40.

«3XXXCnx)

Juvenile stages. Juvenile stages of Solenomeris have often been observed in weakly developed
incrustations, within the basal part of the crust. Only macrospheric forms (A), which are generally
more frequent in all Foraminifera, have been observed (Text. -fig. 5). They show the same
characteristics and the same arrangement of chambers as the Acervulina from the Mururoa atoll
(PI. 2, figs 1-3; Perrin 1985, 1987a). In both cases, juvenile stages are localized close to the encrusted
substratum and differ greatly from the adult stage in size, shape and arrangement of the chambers.
The continuity in growth between juvenile and adult stages refutes the interpretation of a free living
Foraminifera enclosed by acervulinids.

The equatorial layer (or equatorial disc) is formed by a planospiral arrangement of spherical or
subspherical chambers of 65 to 70 //m diameter. In axial section, the equatorial disc is limited to
seven to ten large rounded chambers whose diameter decreases towards the periphery (PI. 2, figs

EXPLANATION OF PLATE 2

Figs 1-8. Solenomeris. Figs 1-3. Alaric Mountain, Aude, France, early Eocene. 1, juvenile stages in axial thin
section; UPS Orsay Alaric 15,5; x 95. 2, juvenile stages in axial thin section; UPS Orsay Alaric 13; x 170.
3, juvenile stages in transverse thin section; UPS Orsay Villerouge; x 100. Fig. 4. Isabena Valley, northern
Spain, early Eocene; adult chambers in axial thin section; UPS Orsay Y1 ; x 95. Fig. 5. Alaric Mountain,
Aude, France, early Eocene; transverse thin section through the roofs and floors of adult chambers showing
the pores and a sinuous chamber; UPS Orsay Alaric 15.3; x 195. Fig. 6. Isabena Valley, northern Spain,
early Eocene; axial thin section of adult chambers; mud infill aids the observation of roof pores in axial
section; UPS Orsay Y1 ; x 1 15. Figs 7-8. Albas, Aude, France, early Eocene. 7, SEM of subaxial section of
adult chambers filled by a sparitic cement; UPS Orsay S2'; x 350. 8, SEM of oblique section of adult
chambers; UPS Orsay S3'; x 350.

UPS, Universite Paris XI.

PLATE 2

PERRIN, Solenomeris

444

PALAEONTOLOGY, VOLUME 37

1-2). Transverse sections are very rare and show the spiral arrangement of the equatorial chambers
(PI. 2, fig. 3).

The lateral chambers have the same shape and size as the adult chambers and form in axial
section a free bipolar growing ovoid as in Acervulina inhaerens.

Adult stage. The adult stage of Solenomeris is encrusting. The geometry and arrangement of adult
chambers (Text-fig. 6) are similar to Acervulina inhaerens. In axial section, this kind of arrangement
forms juxtaposed stacks of chambers, which are more or less perpendicular to the substratum and
have been described by Douville as ‘files of cells’ (1924, pp. 169-170) (PI. 2, figs 4, 7-8).

In axial section, chamber height is c. 15-20 /tm, while average width is 50-60 //m. Some isolated
chambers are, however, larger but seem to be irregularly distributed.

In tangential section, the adult chambers have a rounded and irregular shape. Their average
diameter is 60 //m (from 40 to 120 /im). Some chambers, having a much larger size, show very
irregular, elongated sinuous shapes, their distribution being relatively random (PI. 2, fig. 5).

Because encrustation is often more or less irregular, the growth direction of successive layers can
change rapidly. This results in the gradual change from an axial to a tangential section through
oblique sections which show more or less arched, typical chambers (PI. 3, fig. 1). This vertical
succession of differently oriented sections can often be observed in thin section and may give the
illusion of a change in growth rate of the test. Thus, Moussavian (1989) described the axial and
oblique sections as corresponding to two different kind of chambers which he misinterpreted, like
Douville ( 1924), as the result of slow growth rate ( = axial section) alternating with fast growth rate
(= oblique section). This author also reported the presence of pillars in the external layers of the
test, contributing to its strengthening. These pillars have never been observed in the tests of
Pyrenean Solenomeris. On the other hand, calcitic cementation within the chambers is frequently
observed both in Solenomeris and modern acervulinids and produces a thickening of the walls
which, in axial sections, may be mistaken for the pillars described by Moussavian (1989).

The chamber walls, 5-7 jum thick, are characteristic of hyaline Foraminifera and are composed
of two layers of fibrous hyaline calcite developed on both sides of a dark median layer.

The pores, 5-7 jum in diameter, are clearly visible in tangential sections through the roof of the
chambers (PI. 2, fig. 5) but hardly noticeable in axial section, except when chambers have been filled
by carbonate mud (PI. 2, fig. 6).

In ultrathin sections and in scanning electronic microscopy, the dark median layer of the
tangential walls of Recent Acervulina appears to be continuous through the pores. In the Eocene
Solenomeris , the dark layer is not preserved but its median location can be easily recognized and
appears in scanning electron microscopy as a planar continuous void between the internal and
external hyaline calcite layers.

The lateral walls are not perforated but the presence of stolons has been observed.

text-fig. 6. Adult stage of Solenomeris showing
three-dimensional arrangement of chambers and
microstructure of the tangential wall (after Perrin
1987a); x 400.

PERRIN: ACERVULINID FORAMINIFERA

445

table 12. Size of main features of the test of species of Solenomeris.

Test

Juvenile stages

Adult stage

Thick

ness

(mm)

Ovoid

Prolo-

culus

Equatorial

chambers

Lateral

chambers

References and
localities

0

(mm)

0

C«m)

Height

Cam)

0

(/an)

Width

(/on

Height

Cam)

Width

Cum)

Height

Cam)

Trauth 1918

Polytrema planum

F6

0-36

—

—

—

—

—

60

—

Austria

4-7

1-76

—

—

—

—

—

80

—

Douville 1924

S. O'Gormani

—

—

—

—

—

—

—

35

—

S. France

—

—

—

—

—

—

—

50

—

Pfender 1926

S. douvillei N. Spain

—

—

—

—

—

—

—

30-45

20

S. sp. N. Italy

—

—

—

—

—

—

—

35-60

20-30

Maslov 1956

S. afonensis

—

—

—

—

—

—

—

25-50

10-30

Abkhazie

Rao and Varma 1953

S. douvillei n. Pfender

—

003

280

72-8

—

31

31

—

13

Pakistan

—

F7

412

115-5

—

78

78

—

—

Elliott 1964

S. ogormani

—

—

—

—

—

—

—

40

26

Iraq

—

—

—

—

—

—

—

65

—

Boulanger and Poignant
1969

S. ogormani

—

—

600

200

—

70

70

45

45

S. France

Poignant and Du Chaffaut

—

950

270

—

100

100

50

50

1970

S. ogormani

—

—

420

180

—

40

40

80

35

Corsica France

—

—

—

—

—

—

—

100?

50

Poignant and Blanc 1974

S. ogormani

—

—

600

200

—

70

70

40

—

S. France

—

—

950

270

—

100

100

50

—

Poisson and Poignant
1974

S. douvillei

—

—

—

—

—

—

—

45

20

Turkey

—

—

—

—

—

—

—

50

25

Perrin 1987 a

Solenomeris

—

—

—

—

80

80

80

40

10

S. France N. Spain

—

—

—

—

—

—

—

100

20

Other species

Three other species of Solenomeris have been described (Table 12): S. douvillei Pfender, 1926; S. (?) douvillei
Rao and Varma, 1953 (non Pfender) later called S. pakistense by Johnson and Konishi (1960), and S. afonensis
Maslov, 1956.

Pfender (1926) separated her Solenomeris douvillei from S. ogormani by its encrusting habit, which never
forms autonomous masses, and also reported a more zoned aspect of the test which corresponds in fact to a

446

PALAEONTOLOGY, VOLUME 37

succession of different sections (axial and oblique) and a development which seems a little bit different (Pfender
1926). The 'cell’ size of S. douvillei is similar to that of S. ogormani : 30-45 pm in width and 20 /mm in height.

Solenomeris piae, mentioned by some authors (Boulanger and Poignant 1969; Poignant and Blanc 1974),
corresponds to a specimen from the Upper Cretaceous (?) of Cuba first described by Keijzer (1945) as
Solenopora piae. In accordance with the photographs of Keijzer (1945), this organism appears to be a true
solenoporacean alga and not a Solenomeris.

Rao and Varma (1953) described a new species from the Eocene of Pakistan which differs from Solenomeris
ogormani by its encrusting habit and by a smaller size of its ‘cells’. However, Rao and Varma (1953) compared
the width of the ‘cells’ of S. gormani (i.e. 35-50 pm, according to Douville 1924) with the height of the ‘cells’
measured in their specimens (i.e. 13 /mi and 1 8 — 40 /im). Moreover, these authors distinguished some larger
‘cells’ (18-40 /mi high) in the internal zone of the ‘ thallus’ and smaller ‘cells’ (13 pm high) in the external zone.
Without doubt, both types of ‘cells’ correspond in fact to differently oriented chamber sections (oblique
sections for larger ‘cells’ and axial sections for smaller ‘cells’). Rao and Varma (1953) named this species
5. (?) douvillei , a name preoccupied by S. douvillei Pfender, 1926 of which they seemed unaware. For this
reason, Johnson and Konishi (1960) proposed the new name Solenomeris pakistense.

Lastly, the species S. afonensis was created by Maslov (1956) from Lower Eocene specimens of Novyj Afon
in Abkhazie, but later Elliott (1964) and Poignant (Poisson and Poignant 1974) considered this species identical
to S’. douvillei. The width of the chambers ranges between 25 and 50 pm, while their height is between 10 and
30 pm. The thickness of the wall is c. 10 /mi.

Generic characteristics of Solenomeris

The juvenile stages of Solenomeris show the development of a three-zoned free-living ovoid. During
the earliest stages, the equatorial disc is characterized by the planospiral arrangement of
subspherical equatorial chambers around the proloculus and the second periembryonar chamber
with an enlarged protochonchal spiral (chambers decreasing in size). Afterwards, the ovoid is
formed by addition of lateral chambers around this median equatorial disc.

The adult stage is encrusting and shows a unipolar growth on the dorsal side, characterized by
the formation of successive layers of chambers alternating from one layer to the next one.

In axial section, adult chambers are subhexagonal, while in tangential section they have a
rounded or more rarely sinuous elongated shape.

From the diagnosis given by different authors, the criteria of distinction at the specific level
appear to be based first, on the encrusting or nodular habit and second, on the size of the chambers
(so-called cells) (Tables 12-13).

Growth pattern

Encrustation. The attached stage of Solenomeris , as in other acervulinids, begins with close
encrustation of a rigid substratum, developing with unidirectional formation of successive layers of
chambers (PI. 3, fig. 2). The thickness of such a crust varies from a few millimetres to several
centimetres.

Development of branching form. In Solenomeris , contrary to other acervulinids, this encrusting
growth may be followed by the development of branches, 1-2 centimetres in diameter. These

EXPLANATION OF PLATE 3

Figs 1-5. Solenomeris. Fig. I Isabena Valley, northern Spain, early Eocene; transition from tangential thin
section (lower part of the photograph) to an axial section (top of the photograph); UPS Orsay Y1 ; x 35.
Figs 2-3. Alaric Mountain, Aude, France, early Eocene. 2, early stage of an irregular crust; at the base of
the crust, Solenomeris is interlayered with thin coralline crusts; UPS Orsay Sol. Alaric; x 5-5. 3, thin section
of a branch; UPS Orsay Sol. Alaric; x 6. Figs 4-5. Corbieres, Aude, France, early Eocene. 4, early stage of
the branch development in thin section; the substratum consists of a coral fragment; UPS Orsay R15; x 6.
5, early stage of the branch development in thin section, UPS Orsay R15; x 5-5.

UPS, Universite Paris XI.

PLATE 3

PERRIN, Solenomeris

448

PALAEONTOLOGY. VOLUME 37

table 13. Characteristics of species of Solenomeris described in the literature.

Species

Habit

Size of ‘cells’

Comparisons

S. O' Gorman i [sic]

Nodule

35-50 pm

Type species

Douville 1924

S. douvillei

Crust

30-45 pm width

More zoned than

Pfender 1926

20 /an height

S. 0'Gormani\ no
individualized nodules;
growth slightly different

S. piae Keijzer 1945

7

90- 1 00 pm

‘Cells' are twice as large as
than in other species

S? douvillei non Pfender

Crust

External zone of the test:

Differs from S. ogormani in

Rao & Varma 1953

13 pm height internal zone:

the size of its ‘cells’ and in

= S. pakistense

1 8^40 pm height

its encrusting habit

Johnson and Konishi 1960

S. afonensis
Maslov 1956

Crust

45-50 pm width
20-25 pm height

text-fig. 7. Vertical section of a Soleno-
meris branch showing growth from highly
convex-upwards cupolas at the top of the
branch (traced from a microphotograph
of thin section); c. x 3.

closely-packed branches give buildups a massive aspect (PI. 4, figs 1-3). In longitudinal thin
sections, the banded structure made by the successive layers of chambers shows very convex domes
covering the entire extremities of the branches and contributing at the same time to their thickening
and their lengthening (PI. 3, fig. 3; Text-fig. 7). Thus, the transverse sections of branches are
characterized by tangential sections in their central part changing gradually towards the periphery
to axial sections through intermediate oblique sections.

The development of branches is initially controlled by the morphology of the substratum, the
convex irregularities of which influence the location of branch initiation (PI. 3, figs 2, 4-5). On the
other hand, the branches, which show a typical negative geotropism, can only develop if the
substratum is stable, either as an originally stable substratum or a nucleus secondarily stabilized by

PERRIN: ACERVULINID FORAMINIFERA

449

the weight of the thick Solenomeris crust. Thus, nodules (macroids sensu Hottinger 1983) resulting
from the encrustation of a bioclast by Solenomeris , once stabilized, show a lower smooth surface,
whereas the upper surface shows a progressive development of branches from initial swellings
(Perrin 1992; Plaziat and Perrin 1992).

Bioherms. The buildups constructed by Solenomeris are massive domes with a metre-sized height
and a more or less convex upward shape. These bioherms are made by the coalescent vertical
branches of Solenomeris , with an oblique or fan-like growth at the edges of the buildups (PI. 4, figs
3, 5) (Plaziat 1984).

Biostromes. A cluster of adjacent bioherms constitutes a reef, which is a typical biostrome. The
spaces between the metric domal buildups is filled with a more or less sandy carbonate mud or with
gravels consisting of broken branches of Solenomeris (PI. 4, figs 4, 6). In the latter case,
discrimination between the buildup and detrital parts is especially difficult in weathered outcrops.
These biostromes can extend more than ten kilometres in length, with a thickness which can exceed
ten metres (Plaziat 1984; Plaziat and Perrin 1992).

DISCUSSION

Identification of genera

The systematics of the Acervulinidae is relatively confused, especially at the generic level. Although
Hanzawa (1931, 1947, 1957) and Bursch (1947) established precise specific criteria for the various
species of Acervulina and Gypsina , discrimination between these genera does not seem to be really
based on generic criteria. According to the definitions of Cushman (1950) and Loeblich and Tappan
(1964), the diversity of size, shape and arrangement of the chambers would characterize the genus
Acervulina ; on the other hand, only Gypsina can include free-living spherical species. It seems
difficult to base the distinction on the size of the lateral chambers and the irregularity of their
arrangement. While the spherical shape of Gypsina globulus may be a specific characteristic, the
shape of the test in the other acervulimds is too variable, especially in encrusting forms, to be used
for the identification of the different genera. However, Gypsina appears to be represented by species
of smaller size made of a limited number of layers of chambers.

According to Hanzawa (1931, 1940, 1947, 1957) and Bursch (1947), the test of acervulinids
appears to consist of three clearly different zones which can be recognized in axial sections: a
median layer or equatorial zone, formed by the equatorial disk during the juvenile stages, which is
enclosed between a dorsal zone and a ventral zone. Some of these three zones may be absent in some
species of Gypsina. On the other hand, all three are always present in Borodinia and Acervulina. The
number of layers of chambers within the ventral and dorsal zones changes in accordance with the
different forms of acervulinids. Only the equatorial zone, when it exists, is constantly represented
by a single layer of chambers. The number of distinct zones constituting the test, and the number
of layers within the ventral and dorsal zones, appear to be the diagnostic criteria distinguishing
Gypsina from Acervulina and Borodinia.

A third criterion seems especially reliable for generic discrimination. This concerns the thickness
of the tangential walls (roofs and floors of the chambers) which is relatively constant within a genus.
Thus, Acervulina and Gypsina are both characterized by thin tangential walls, whose thickness varies
between a few microns and 25 pm. On the other hand, the thickness of the roofs in Borodinia is more
than 100 /an (140 pm in the species described by Hanzawa 1957).

A combination of these three criteria provides a relatively easy and reliable method for separating
the genera in the Acervulinidae (Table 14). Acervulina is only represented by encrusting forms with
a three-zoned test. The dorsal zone always comprises several layers of chambers, whose number is
much higher than that of the ventral zone. The tangential walls are thin (5-25 //m). Borodinia
possesses three distinct zones, the dorsal one always including several layers of chambers. The
tangential walls are thicker than 100 /nn. Gypsina includes both free and encrusting forms with a

450

PALAEONTOLOGY, VOLUME 37

test consisting of one, two or three distinct zones. The dorsal zone, when it exists, is formed by one
single, or a few layers of chambers, the number of dorsal layers always remaining lower or
equivalent to that of the ventral zone. The tangential wall of chambers is thin (5-25 //m).

The use of these criteria requires observation of the test in axial sections passing through the
ovoid of juvenile stages in order to distinguish the different zones. It is not necessary to have a
section of the proloculus for the identification of acervulinids at the generic level. However, a precise
measurement of the tangential walls must be taken in axial section and not in oblique sections. This
can be made difficult by diagenesis, especially when the chambers have been cemented by fibrous
calcite which it is critical to distinguish from the fibrous walls of the chambers. Nevertheless, the
difference in thickness between the thin tangential walls (5-25 //m) of Gypsina and Acervulitia, and
the thick walls of Borodinia (more than 100//m) is important enough to constitute an especially
reliable and easy to use criterion of identification at generic level.

Identification of species

Species of Gypsina may be easily distinguished according to the number of zones (ventral, median
or equatorial, and dorsal), and the number of layers of chambers within each of these zones.

Borodinia is only represented by a single species: Borodinia septentrionalis described by Hanzawa
(1940, 1957). The presence of the three zones constituting the test and the larger thickness of the
walls are criteria characterizing the genus. The species is distinguished by spatuliform chambers in
tangential sections (Hanzawa 1957).

In Acervulina , the criteria which may be used for species identification are the morphology of the
roof of the chambers in axial section and the shape of some chambers in tangential section.
A. linearis is characterized by roofs of the lateral chambers of the same layer forming a continuous
straight line in axial section. However, in the other species, the adult chambers have roofs which are
separate from each other. In tangential section the lateral chambers of A. linearis have angular
shapes while those of A. inhaerens are rounded. Acervulina ( Ladoronia ) vermicularis shows large
sinuous chambers in tangential sections.

Validity of species of Solenomeris

The distinction between different species of Solenomeris is based on two criteria: growth habit and
size of the adult chambers ('cells’) (Table 13). The growth forms, as in many sessile organisms,
depend directly on environmental factors: morphology of the substratum, hydrodynamics and
competition with other organisms. Their morphology also changes during the development of an
individual. Branching Solenomeris necessarily begins after an encrusting stage. In the same way,
macroids from the Corbieres show a primary concentric encrustation stage and, after the
stabilization of the nodule by its own weight, begin to develop branches on the upper side (Perrin
1992; Plaziat and Perrin 1992). Therefore, the growth habit of the organism does not appear to be
a reliable criterion for the identification of species.

EXPLANATION OF PLATE 4

Figs 1-6. Solenomeris. Figs 1-3. Alaric Mountain, Aude, France, early Eocene. 1, outcrop view showing the
closely packed branches of Solenomeris on a weathered surface. 2, branches in growth position; note the
presence of scarce muddy sediment between branches (photograph width is c. 30 cm). 3, detail of a
Solenomeris bioherm showing bivalve borings filled with muddy sediment, (photograph width is c. 35 cm).
Fig. 4. Albas reef, Aude, France, early Eocene; fragments of branches in a muddy sediment. Figs 5-6. Alaric
Mountain. Aude, France, early Eocene. 5. detail of a Solenomeris bioherm showing the vertical or oblique
growth of branches. 6, part of a Solenomeris reef showing the cluster of adjacent bioherms and locally the
muddy infill between the bioherms (photograph width is c. 4 m).

UPS, Universite Paris XI.

PLATE 4

PERRIN, Solenomeris

452

PALAEONTOLOGY, VOLUME 37

The dimensions of the chambers, diagnostic data for palaeontologists treating Solenomeris as an
alga, are difficult to compare when axial and transverse sections have not been distinguished. On
the other hand, authors considering Solenomeris as a foraminifer do not specify the size of the
chambers (Hagn 1967; Hagn and Wellnhofer 1967; Moussavian 1989). Thus, the heights of the
chambers in different Solenomeris descriptions are especially difficult to interpret since measure-
ments may have been taken in oblique sections and are therefore overestimated. This is certainly the
case for the sizes reported by Rao and Varma (1953) who distinguished between smaller sized ‘cells’,
located in the external part of the test, and larger ‘cells’ situated in the internal zone (Table 13).
These correspond respectively to axial sections and oblique sections of the same type of chambers.
Without taking into account the measurements of the larger chambers reported by Rao and Varma
(1953), the height of chambers appears to vary between 13 pm in S. pakistense (Rao and Varma
1953) and 30 /nn in Solenomeris from the Middle Eocene of northern Italy (Pfender 1926). In the
studied samples from the Early Ypresian of the Corbieres and northern Spain, average height
varies from 15 to 20 //m and ranges between 10 and 35 pm. However, even if previous authors gave
overestimated dimensions, the variability in height of chambers in all the so-called species of
Solenomeris shows the same range of size as the variability which exists at the specific level in
Acervulina and Gypsina. For example, the height of chambers can vary from 5 to 25 //m in
G. vesicular is (Bursch 1947; Hanzawa 1957), from 5 to 47 /mi in A. inhaerens (Hanzawa 1931, 1957),
and from 11 to 50 //m in A. linearis (Hanzawa 1947, 1957) (Tables 1-2, 5).

On the other hand, the width of the chambers seems to be quite similar in the four species
Solenomeris ogormani , S. douvillei , S. pakistense and S. afonensis and varies from 25 to 50 //m.
Solenomeris from southern France and northern Spain have lateral chambers whose size varies from
40 to 100 //m, with an average of 60 /mi. Moreover, some tangential sections show the presence of
sinuous larger chambers whose length can reach 250-300 pm. However, these large-sized chambers
are relatively rare and are seldom reported in previous descriptions of Solenomeris (Perrin 1985,
1987u). In Acervulina and Gypsina , the variability of chamber width within the same species is
important, especially in G. globulus (60-180 pm) and in A. inhaerens (50-230 //m).

Therefore the size of the adult lateral chambers does not seem to be a reliable criterion for
distinguishing different species of Solenomeris because of the large intraspecific variability of this
feature in the acervulinids. Some authors have synonymized the different species of Solenomeris but
without any discussion (Elliott 1964; Boulanger and Poignant 1969; Poignant and Blanc 1974;
Moussavian 1989).

Comparison between Solenomeris and other Acervulinidae

The test of Solenomeris is typically composed of three distinct zones: a ventral zone formed by the
lateral chambers of the ovoid, juvenile stages; a median or equatorial zone comprising a single layer
of chambers which constitutes the equatorial disc including the proloculus; and a dorsal zone made
of numerous layers consisting of lateral chambers of the juvenile stages and chambers of the adult
stage which constitute the bulk of the construction.

The tangential walls of Solenomeris are thin (5-10 //m). These characteristics of Solenomeris
appear closer to the genus Acervulina than to any other member of the Acervulinidae.

In axial section, the roofs of the lateral chambers in Solenomeris never form a continuous line like
A. linearis , but appear clearly individualized (when skeletal preservation is good enough).
Tangential sections mainly show rounded chambers but with scarce larger and sinuous chambers
similar to those of A. vermicularis. Thus, Solenomeris ogormani differs both from Acervulina
inhaerens and A. ( Ladoronia ) vermicularis since its test comprises, in tangential section, not only
rounded lateral chambers but also large sinuous chambers (Table 14).

Moussavian (1989) united the different species of Solenomeris as Acervulina ogormani instead of
Gypsina ogormani (Moussavian 1984) because he believed that the genera Acervulina Schultze, 1854
and Gypsina Carter, 1877 were synonymous, Gypsina being the junior synonym. Nevertheless,
Moussavian suggested that an extensive revision of this family was necessary. The synonymy of
Acervulina and Gypsina is also based on the long- recognized synonymy of the species A. inhaerens

PERRIN: ACERVULINID FORAMINIFERA

453

table 14. Key to main genera and species of Acervulinidae.

A Test free or encrusting, consisting of one, two or three distinct zones; dorsal zone, when present,
comprising one or a few layers of chambers; thin tangential walls (5-25 //m)

Gypsina

A1 One single zone

• Equatorial zone: one single layer G. squamiformis

• Equatorial zone absent G. globulus

A2 Two zones

• Equatorial zone: one single layer of chambers and dorsal

zone G. mastelensis

A3 Three zones

• Equatorial zone: one single layer of chambers

Ventral and dorsal zones: equal number of layers G. vesicularis

• Equatorial zone: one single layer of chambers
Ventral zone: a few chambers

Dorsal zone: one single layer G. saipanensis

• Equatorial zone: one single layer of chambers
Ventral zone: several layers of chambers

Dorsal zone: two or three layers G. marianensis

B Test encrusting, consisting of three distinct zones: a ventral zone,
an equatorial zone and a dorsal zone; dorsal zone comprising several
layers of chambers

B1 Thick tangential walls (> 100 /on)

• Some chambers showing a spatulifornr shape in tangential
section

B2 Thin tangential walls (5-25 /nn); juvenile stages with equatorial
chambers increasing in size towards the periphery of the
equatorial disc

• Roofs of chambers of the same layer forming a continuous
straight line in axial section; chambers more or less angular
in tangential section

• Roofs of chambers of the same layer do not form a
continuous straight line in axial section; tangential section
showing only rounded-shaped chambers

• Roofs of chambers of the same layer do not form a
continuous straight line in axial section; tangential section
showing sinuous large chambers

B3 Thin tangential walls (5-25 /nn); juvenile stages with equatorial
chambers decreasing in size towards the periphery of the
equatorial disc

• Roofs of chambers of the same layer do not form a
continuous straight line in axial section; most of the
chambers of tangential section show a rounded shape but
scarce sinuous large chambers are also present

Borodinia
Borodinia septentvionalis

A cervulina

Acervulina line avis

A. inhaerens

A. ( Ladovotiia) vevmiculavis

Solenomevis

Solenomevis ogovmani

and G. plana. Hanzawa (1957) considered G. plana as a variety of A. inhaerens : Acervulina inhaerens
var. plana. However, the type species of Gypsina is not Polytrema planum = Gypsina plana Carter,
1877, as indicated by Loeblich and Tappan (1964), but Gypsina vesicularis Parker and Jones, 1860.
Moreover, from the above review of the geometrical characteristics of the test, both Acervulina and
Gypsina appear clearly different and can be easily distinguished by the number of zones constituting

454

PALAEONTOLOGY, VOLUME 37

the test and by the number of layers of chambers within the dorsal and ventral zones. It is therefore
difficult to place the different species of Solenomeris within any particular genus without first
carefully studying the geometrical features of their internal structure.

CONCLUSIONS

The foregoing critical review of growth organization and chamber morphologies in Acervulina,
Borodinia , Gypsina and Solenomeris suggests reliable and easily identifiable criteria for the
discrimination of different genera and species (Table 14). The genus Solenomeris , recently included
in the Acervulinidae (Hagn 1967; Perrin 1985, 1987a), is certainly closely related to the genus
Acervulina. However, the morphology of the roof of the chambers in axial section, the occurrence
of numerous rounded lateral chambers and also of large sinuous chambers distinguish clearly
Solenomeris from species of Acervulina. Thus also taking into account the form of the juvenile with
its enlarged protoconchal spiral, it would seem appropriate to regard Solenomeris as a separate
genus within the Acervulinidae.

Previously described species of Solenomeris seem to be identical because their discrimination is
based on unreliable criteria; their geometry and the internal structure of the test having been
misinterpreted as those of a red alga.

Acknowledgements. I thank J. C. Plaziat (Universite Paris-Sud, Orsay) for many discussions which helped in
improving the manuscript. Many thanks are also expressed to B. R. Rosen and J. E. Whittaker (Natural
History Museum, London) for their useful comments on the manuscript.

REFERENCES

barbin, v., ramseyer. k., decrouez, d. and herb, r. 1989. Mise en evidence par la cathodoluminescence
d’indices de remaniements synsedimentaires. Geobios , 22, 253-259.
beckman, j. p.. Born, h. m., kleboth, p. and decima, f. p. 1982. Micropaleontology and biostratigraphy of the
Campanian to Paleocene of the Monte Giglio, Bergamo Province, Italy. Memorie di Scienze Geologiche , 35,
91-172.

bermudez, p. j. 1952. Estudio sistematico de los Foraminiferos rotaliformes. Boletin de Geologia , 2, 1-230.
betzler, c. 1987. History of a Paleocene carbonate platform in a foreland basin: the Sierra del Port del Comte
and the Sierra del Cadi (Southern Pyrenees, NE Spain). Abstracts, 8th International Association of
Sedimentologists Regional Meeting of Sedimentology, Tunis 1987, Poster Session , 96-97 .

1989. The upper Paleocene to middle Eocene between the Rio Segre and the Rio Llobregat (Eastern South
Pyrenees): facies, stratigraphy and structural evolution. PhD Thesis, Tiibinger Geowissenschaftliche
Arbeiten, A (2), 1 13 pp.

boulanger, d. and poignant, a. f. 1969. Sur les nodules algaires du Lutetien superieur de Sainte-Marie-de-
Gosse (Landes). Comptes Rendus Sommaires de la Societe Geologique de France, 4, 109.
brady, h. b. 1884. Report on the Foraminifera dredged by H.M.S. Challenger during the years 1873-1876.

Scientific Results of the Voyage of H.M.S. Challenger during the years 1873-1876 , Zoology, 9, 1-814.
brugnatti, m. a. and ungaro, s. 1987. Analogie e differenze tra Solenomeris (Alga) e Gypsina (Foraminifero).

Annali delT Universitci di Ferrara, Sezione 9, Scienze Geologiche e Paleontologiche, 9, 1-14.
bursch, j. g. 1947. Mikropalaontologische Untersuchungen des Tertiar von Gross Kei (Molukken).

Schweizerische Paldontologische Abhandlungen, 65, 1-169.
carpenter, w. b. 1876. Remarks on the Carter’s paper ‘On the Polytremata (Foraminifera), especially with
reference of their hybrid nature’. Annals and Magazine of Natural History, (4), 17, 380-387.
carter, h. j. 1876. On the Polytremata (Foraminifera), especially with reference of their hybrid nature. Annals
and Magazine of Natural History, (4), 17, 185-214.

1877. On a Melobesian form of Foraminifera (Gypsina melobesioides, mild) ', and further observations on
Carpenteria monticularis. Annals and Magazine of Natural History, (4), 20, 172- 176.
chapman, f. 1900. Foraminifera from the lagoon at Funafuti. Journal of the Linnean Societv (Zoology), 28,
I61 -2 1 0.

cushman, j. a. 1915. Monograph of the Foraminifera of the North Pacific Ocean. Bulletin of the United States
National Museum, 71, 1-596.

PERRIN: ACERVULINID FORAMINIFERA

455

— 1950. Foraminifera, their classification and economic use. Fourth Edition. Flarvard University Press,
Cambridge, 605 pp.

— todd, r. and post, r. j. 1954. Recent Foraminifera of the Marshall Islands, Bikini and nearby atolls; part
2, Oceanography (Biologic). United States Geological Survey Professional Papers , 260-H, 319-384.

deloffre, R., poignant, a. f. and TEHERANi, k. 1977. Algues Calcaires de l’Albo-Aptien au Paleocene de l’lran
Central. Bulletin des Centres de Recherche Exploration- Production Elf- Aquitaine, 1, 29-57.
de zanche, v. 1965. Le microfacies eoceniche nella Valle del Chiampo tra Arzignano e mussolino (Lessini
oricntali). Rivista Italiana di Paleontologia e Stratigrafiya , 71, 925-948.

— sorbini, l. and spagna, v. 1977. Geologia del Territorio del comune di Verona. Memorie del Mitseo Civico
di Storia Naturale di Verona , (2), Sezione Scienze della Terra , 1, 1 51.

dieni, t, massari, f. and poignant, a. f. 1979. Testimonianze di Paleocene marino in Sardegna. Rivista Italiana
di Paleontologia e Stratigrafiya , 85, 481-516.

douville, h. 1924. Un nouveau genre d’Algues calcaires. Comptes Rendus Sommaires de la Societe Geologique
de France , 4, 169-170.

— and o’gorman, g. 1924. L’Eocene inferieur du Bearn. Comptes Rendus Sommaires de la Societe
Geologique de France . 4, 170-172.

dullo, w. c., moussavian, e. and brachert, t. c. 1990. The foralgal crust facies of the deeper fore reefs in the
Red Sea: a deep diving survey by submersible. Geobios , 23, 261 281.
eichenseer, h. 1987. Early Eocene carbonate platforms with reefs at the northern flank of the South Pyrenean
foreland basin (NE Spain). Abstracts , 8th International Association of Sedimentologists Regional Meeting of
Sedimentology , Tunis 1987 , 217-218.

— 1988. Facies geology of late Maastrichtian to early Eocene coastal and shallow marine sediments (Tremp-
Graus Basin, Northeastern Spain). Ph D. Arbeiten aus dem Institut und Museum fur Geologie und
Palaontologie der Universitat Tiibingen, 237 pp.

elliott, g. f. 1960. Fossil calcareous algal floras of the Middle East with a note on a Cretaceous
problematicum Hensonnella cylindrica gen. et sp. nov. Quaterly Journal of the Geological Society , London,
115, 217-232.

— 1964. Tertiary solenoporacean algae and the reproductive structures of the Solenoporaceae.
Palaeontology, 7, 695- 702.

franzenau, a. 1893. Semseya, eine neue Gattung aus der Ordnung der Foraminiferen Mathematische und
Naturwissenschaftliche Berichte aus Ungarn, 11, 358-361.
gaemers, p. a. m. 1978. Biostratigraphy, palaeoecology and palaeogeography of the mainly marine Ager
Formation (upper Paleocene-lower Eocene) in the Temp basin. Central - South Pyrenees, Spain. Leidse
Geologische Mededelingen, 51, 151-231.

galloway, j. j., 1933. A manual of Foraminifera. James Furman Kemp Memorial Series Publication /,
Principia Press, Bloomington, 483 pp.

— and wissler, s. G. 1927. Pleistocene Foraminifera from the Lornita Quarry, Palos Verdes Hills,
California. Journal of Paleontology, 1. 35-87.

garavello, a. m. and ungaro, s. 1982. Studio biostratigrafico e paleoecologico della serie eocenica di
Pradipaldo nella zona pedemontana meridionale dell’ altipiano di Asiago (Vicenza). Geologica Romana, 21,
655-675.

goes, a. 1882. On the reticularian Rhizopoda of the Caribbean Sea. Handlingar Konglica Svenska
VetenskapsAkademiens Bihang , 19 (4), 1-151.

hagn, h. 1967. Das Alttertiar der bayerischen Alpen und ihres Vorlandes. Mitteilungen Bayerischen
Staatsammlung fur Palciontologie und Historische Geologie, 7, 245-320.

— 1972. Uber kalkalpme paleozane und untereozane Gerolle aus dem bayerischen Alpenvorland.
Mitteilungen Bayerischen Staatsammlung fur Palciontologie und Historische Geologie, 12, 113-124.

— 1978. Die alteste Molasse im Chiemgau/ostliches Oberbayern (Katzenloch-Schichten, Priabon).
Mitteilungen Bayerischen Staatsammlung fur Palaontologie und Historische Geologie, 18, 167-235.

1983. Die Kreide- und alttertiar-Gerolle des Wachtberg-Schotters (Ottnang Subalpine Molasse)
N Salzburg. Mitteilungen Bayerischen Staatsammlung fur Palciontologie und Historische Geologie, 23,
125-150.

— and moussavian, e. 1980. Die Gosau- und Alttertiargerolle des Westerbuchbergs (Unt. Eger, subalpine
Molasse, Chiemgau). Mitteilungen Bayerischen Staatsammlung fur Palaontologie und Historische Geologic,
20, 137- 157.

and wellnhofer, p. 1967. Ein errastiches Vorkommen von kalkalpinen Obereozan in Pfaffing bei

Wasserburg. Geologica Bavarica, 57, 205-288.

456

PALAEONTOLOGY, VOLUME 37

hansen, h. j. and reiss, z. 1971. Electron microscopy of rotaliacean wall structures. Bulletin of the Geological
Society of Denmark, 2014. 329-346.

hanzawa, s. 1931. Notes on Tertiary foraminiferous rocks from the Kwanto Mountainland, Japan. Scientific
Reports of the Tohoku Imperial University , 2, {Geology), 12 (2A), 141-157.

— 1939. On the occurrence of Acervulina, an encrusting form of Foraminifera in the Jurassic Torinosu
limestone of the Kwanto Mountainland, Central Japan. Journal of the Geological Society of Japan, 46 (547),
201-203.

— 1940. Micropaleontological studies of drill cores from a deep well in Kita-Daito-Zima (North Borodino
Island). 755-802. In yabe h. (ed. ). Jubilee Publication in the Commemoration of Professor H. Yabe's 60th
Birthday. Tokyo, 1251 pp.

1947. Note on an Eocene foraminiferal limestone from New Britain. Japanese Journal of Geology and
Geography, 20, 59-6 1 .

— 1957. Cenozoic Foraminifera of Micronesia. Geological Society of America, Memoir , 66, 1-163.
haynes, J. R. 1981. Foraminifera. J. Wiley and Sons, New York, 433 pp.

hottinger, l. 1983. Neritic macroid genesis, an ecological approach. 38-55. In peryt, t. (ed.). Coated grains.
Springer-Verlag, Berlin, 655 pp.

Johnson, j. h. and konishi, k. 1960. An interesting late Cretaceous calcareous alga from Guatemala. Journal
of Paleontology, 34, 1099-1105.

keijzer, f. j. 1945. Outline of the geology of the eastern part of the Province of Oriente, Cuba. Geographische
Geologische Mededelingen , 2, 1-239.

loeblich, a. r. and tappan, h. 1964. Sarcodina chiefly ‘Trecamoebians’ and Foraminiferida. In moore, r. c.
(ed.). Treatise on invertebrate paleontology. Part C, Protista 2(2). Geological Society of America and
University of Kansas Press, New York and Lawrence, Kansas, 900 pp.

— 1984. Suprageneric classification of the Foraminiferida (Protozoa). Micropaleontology, 30, 1-70.
logan, b. w., harding, J. l., ahr, w. m., williams, j. d. and snead, r. G. 1969. Carbonate sediments and reefs,

Yucatan Shelf, Mexico, Part 2, Coral reefs and banks, Yucatan Shelf, Mexico. Memoir of the American
Association of Petroleum Geologists, 11, 129-198.

maslov, v. p. 1956. Iskopaemye izwest kovye vodorosli SSSR. Trudy Geologicheskikh Institut Akademii Nauk,
SSSR. 160, 1-301. [In Russian; translation in BRGM, Paris, 4934].
massieux, M. 1961 . Facies a Algues de I’Ypresien du bassin d'AIbas (Aude). Comptes Rendus Sommaires de la
Societe Geologique de France , 2, 40-41.

— 1973. Micropaleontologie stratigraphique de l’Eocene des Corbieres Septentrionnales (Aude). Cahiers de
Paleontologie, 1-146.

moussavian, e. 1984. Die Gosau- und Altertiar-Gerolle der Angerberg-Schichten (hoheres Oligozan,
Unterinntal, Nordliche Kalalpen). Facies, 10, 1-86.

— 1989. Uber die systematische Stellung und die Bestimmungskriterien der Solenoporaceen (Rhodo-
phyceae). Courier Forschungs-Institut Senckenberg, 109, 51-91.

orszag-sperber, f., poignant, a. f. and poisson, a. 1977. Paleogeographic significance of rhodolites: some
examples from the Miocene of France and Turkey. 286-294. In flugel, e. (ed.). Fossil Algae, Springer-
Verlag, Berlin, 375 pp.

Parker, w. and jones, T. 1 860. On the nomenclature of Foraminifera. Annals and Magazine of Natural History,
6, 29M0.

perrin, c. 1985. Paleoecologie et diagenese des recifs a Madreporaires et a Solenomeris de I’llerdien d’AIbas
(Aude). Unpublished Memoire Diplome d’Etudes Approfondies, Universite Paris-Sud Orsay, 55 pp.

— 1987m Solenomeris un Foraminifere Acervulinidae constructeur de recifs. Revue de Micropaleontologie,
30, 197-206.

— 1 9876. Role des Foraminiferes encroutants dans l’edification des recifs. Livre des Resumes, ler Congres
Frangais de Sediment ologie, Paris 1987, 274-275.

— 1987c. Les recifs monospecifiques a Foraminiferes de FUerdien des Corbieres. Livre des Resumes,
7" Congres Frangais de Sedimentologie, Paris 1987 , 273.

— 1989. Role des organismes dans 1' edification et I evolution de 1' atoll de Mururoa ( Polynesie Frangaise).
Unpublished Ph.D thesis, Universite Paris-Sud, Orsay.

— 1990. Genese de la morphologie des atolls: le cas de Mururoa (Polynesie Frangaise). Comptes Rendus
de I' Academic des Sciences, Paris, (2), 31 1, 671-678.

— 1992. Signification ecologique des Foraminiferes Acervulinides et leur role dans la formation de facies
recifaux et organogenes depuis le Paleocene. Geobios, 25, 725-751.

— and plaziat, j. c. 1987. Solenomeris , un Foraminifere constructeur dc recifs. Abstracts, 8th International
Association of Sedimentologists Regional Meeting of Sediment ology, Tunis 1987, 404-405.

PERRIN: ACERVULINID FORAMINIFERA

457

pfender, j. 1926. Sur un organisme du Nummulitique de la colline de San Salvador pres Canrmarasa. Boletin
de la Real Sociedad Espanola de Historia Natural , 26, 321 -330.
plaziat, J. c. 1984. Stratigraphie et evolution paleogeographique du domaine pyreneen de la fin du Cretace
(phase Maastrichtienne ) d la fin de F Eocene (phase pyreneenne). Unpublished thesis, Orsay.

— and perrin, c. 1992. Multikilometer-sized reefs built by Foraminifera ( Solenomeris ) from the early
Eocene of the Pyrenean domain (S. France, N. Spain) - paleoecologic relations with coral reefs.
Palaeogeography , Palaeoclimatology, Palaeoecology , 96. 195-231.

poignant, a. F. 1974. Les Algues calcaires fossiles: leur interet stratigraphique. Newsletters on Stratigraphy, 3,
181-192.

1975. Les Algues fossiles: leur role constructeur; leur role d'indicateur d’environnement. Cahiers
Geologiques, 91, 87-100.

1977. Contribution a l’etude des Algues Rouges paleocenes du Bassin Parisien. Bulletin Information des
Geologues du Bassin de Paris , 14, 57-61.

— and du chaffaut, a. s. 1970. Les Algues des formations transgressives maestrichtiennes, paleocenes et
ypresiennes de la cote sud-orientale de la Corse. Revue de Micropaleontologie, 12, 202-208.

— and blanc, p. l. 1974. Les Algues calcaires du Cretace superieur des rides de Villagrains Ladiras et
Roquefort-Creon (Aquitaine Centrale). Geobios , 7, 71-75.

poisson, a. and poignant, a. f. 1974. La formation de Karabayir, base de la transgression miocene dans la
region de Korkuteli (Departement d’Antalya-Turquie). Lithothamnium pseudoramossissimum , nouvelle
espece d'Algue Rouge de la formation de Karabayir. Bulletin of the Mineral Research and Exploration
Institute of Turkey, 82, 67-71.

rao, s. r. n. and varma, c. p. 1953. Fossil algae from the Salt Range, II, Solenomeris (?) douvillei sp. nov. from
the Laki (lower Eocene) limestones. Palaeobotanist, 2, 19-23.
reiss, z. and hottinger, l. 1984. The Gulf of Aqaba : ecological micropaleontology. Ecological Studies, 50.
Springer-Verlag, Berlin, 354 pp.

repellin, p. 1977. Contribution a l’etude petrologique d’un recif corallien: le sondage Colette, atoll de
Mururoa. Cahiers du Pcicifique , 20, 1-210.

reuss, a. 1 848. Die fossilen Polyparien des Wiener Tertiarbeckens ( 1 847). Naturwissenschaftliche Abhandlungen
(Wilhelm Haidinger ), 2. 1-109.

Samuel, o., borza, K. and rohler, e. 1972. Microfauna and lithostratigrapliy of the Paleogene and adjacent
Cretaceous of the Middle Vah Valley ( West Carpathian). Geologicky Ustav Dionyza Slura, Bratislava,
246 pp.

schalekova, a. 1963. Die Algenflora der kretazischen und palaogenen - Kalksteine der Slovakei. Geologicky
Sbornik , 14, 165-167.

SCHULTZE, M. s. 1854. Ueber den Organismus der Polythalamien ( Foraminiferen ), nebst Bemerkungen iiber die
Rhizopoden im Allgemeinen. Wilhelm Engelman, Leipzig, 68 pp.
silvestri, a. 1937. Foraminiferi dell’Oligocene e dell Miocene della Somalia. Paleontographie Italica, 32,
45-264.

tambareau, Y. and villatte, J. 1974. Livret-guide de F excursion dans le Thanetien et Tllerdien de la zone sous-
pyreneenne. Chainons audois , Plantaurel, Petites Pyrenees. Travaux du Laboratoire de Geologie et de
Petrologie de l'Universite Paul Sabatier, Toulouse, 48 pp.
tappan, h. t. 1980. Palaeobiology of plant protists. Freeman, San Francisco. 1028 pp.
terry, c. e. and williams, J. J. 1969. The Idris “A” bioherm and oilfield, Sirte Basin, Libya - its commercial
development, regional Paleocene geologic setting and stratigraphy. 31-48. In hepple, p. (ed.). The
exploration for petroleum in Europe and North Africa. Institute of Petroleum, London, 281 pp.
toumarkine, m. 1966. Etude micropaleontologique et stratigraphique de l'Eocene marin du Mont-Cayla
(Aude) et des ecailles de Saint-Chinian (Herault). Unpublished Thesis, Paris.

1967. Etude micropaleontologique et stratigraphique de l’Eocene marine du Mont-Cayla (Aude) et des
ecailles de Saint-Chinian (Herault). Bulletin de la Societe Geologique de France, 7. 355-366.
trauth, F. 1918. Das Eozanvorkommen bei Radstadt im Pongau und seine Beziehung zu den gleichaltrigen
Ablagerungen bei Kirchberg am Wechsel und Wimpassing am Leithagebirge. Denkschriften der Akademie
der Wissenschaften : Mathematisch-naturwissenschaftliche Klasse Wien. 95, 171-278.
wray, j. l. 1977. Calcareous Algae. Developments in palaeontology and stratigraphy, 4. Elsevier, Amsterdam,
1 85 pp.

yabe, h. and hanzawa, s. 1925. A geological problem concerning the raised coral-reefs of the Riukiu Islands
and Taiwan; a consideration based on the fossil Foraminifera faunas contained in the raised coral-reef

458

PALAEONTOLOGY, VOLUME 37

formation and the youngest deposits underlying it. Scientific Reports of the Tohoku Imperial University , (2),
7, 29-57.

Typescript received 15 January 1993
Revised typescript received 13 September 1993

CHRISTINE PERRIN

Department of Geology
Royal Holloway and Bedford New College
University of London
Egham

Surrey TW20 OEX, UK

A QUANTITATIVE REVIEW OF THE HORSE
EQUUS FROM SOUTH AMERICA

by jose l. prado and maria t. alberdi

Abstract. The species of Equus (Amerhippus) are revised using multivariate analysis of dental and autopodial
remains, and some morphological characters of the skull and dentition. We recognize five species:
E. (A.) andium , insulatus , neogeus , santaeelenae , and lasallei. They have been recorded in Argentina, Bolivia,
Brazil, Colombia and Ecuador, through the Middle and Upper Pleistocene. They show peculiar adaptive
features in the distal part of the extremities as a response to different environmental conditions.

The Quaternary equid record from South America is very rich. From a palaeobiogeographic point
of view, the remains are distributed throughout the continent, from Ecuador to central Argentina.
There are very important quantitative and qualitative collections representing all parts of the
skeleton of all species.

Equus arrived in South America after the Great American Biotic Interchange (Webb 1976). From
Lund (1840) to the present, numerous records and new species have been attributed to this genus.
Sefve (1912), and Boule and Thevenin (1920) have studied South American Equus in monographic
works. Hoffstetter (1952), in his review of Ecuadorian mammals, recognized only six species of this
genus, all included in the subgenus Amerhippus Hoffstetter, 1950. He also included in this South
American subgenus, one North American species Equus occidentals. Porta (1960) described a new
species collected by Brother Daniel, Equus lasallei , from the Late Pleistocene of Colombia. More
recently, MacFadden and Azzaroli (1987) discussed the status of the species cited by Hoffstetter
( 1952) and recognized as valid species: E. andium , E. lasallei , E. insulatus and possibly E. curvidens ,
without a clear definition of the others (E. santaeelenae , E. martinei , E. neogeus). Azzaroli (1992)
provisionally recognized eight valid species from South America (Equus curvidens , E. andium , E.
insulatus , E. haasei , E. martinei , E. santaeelenae and E. lasallei , and the poorly known E. neogeus
with doubts), all ascribed to the subgenus Amerhippus. He also included in this group two North
American species: E. fraternus and E. conversidens.

The subgenus Amerhippus is recorded in South America from the Middle Pleistocene (Ensenadan
Land Mammal Age), appearing after the genus Hippidion , which was first recorded from the Late
Pliocene in that continent. That group of horses is more evolved than Hippidion which has some
primitive characters (Alberdi and Prado 1992, 1993). Species of Amerhippus show two different
morphological patterns; one of them corresponds to forms adapted to grasslands and open
habitats, and the other to mountain forms. Each group is quite homogeneous, with little variation
in size throughout its geographical and stratigraphical distribution. This point has not been given
sufficient consideration in previous taxonomic reviews, which describe large numbers of species.

The aims of this study are: (a) to identify the groups of Equus ( Amerhippus ) resulting from
multivariate analysis of the upper and lower dentitions, metacarpals (MCIII), calcaneum (CA),
astragalus (AS), metatarsals (MTIII) and first phalanges of the third digit (1FALIII); (b) to
determine which morphometric characteristics and bones allow the best identification of the
different species; (c) to redefine the generic and specific diagnoses of South American horses, and
(d) to establish valid taxonomic names.

| Palaeontology, Vol. 37, Part 2, 1994, pp. 459-48 1.|

© The Palaeontological Association

460

PALAEONTOLOGY, VOLUME 37

1 La Carolina
2- Oil Field
3.- Oton
4- Rfo Chiche

5. - Punfn (several sites)

6. - Cerro Gordo

7. - Tarija

8. - Lagoa Santa

9. - Corumba

10 - Sao Raimundo Nonato
1 1 Chique- Chique
12.- Aguas do Araxa
13 - Lujan

14. - Ayacucho

15. - Tapalque
16 - Paso Otero

17. - Necochea

18. - Quequen Salado

19. - Montehermoso

100 km

O Equus {A.) neogeus
☆ Equus {A.) lasallei
□ Equus (A.) insulatus
B Equus (A.) santaeelenae
A Equus (A.) andium

text-fig. I. Geographic distribution of Equus ( Amerhippus ) localities.

MATERIAL AND METHODS

The studied material includes the material studied by Hoffstetter (1952) and Boule and Thevenin
(1920), in the Museo de la Escuela Politecnica Nacional of Quito, Ecuador (MEPN), Institut de
Paleontologie du Museum National d'Histoire Naturelle of Paris, France (IPMNHN) and Museo
Nacional de Elistoria Natural (MNHN) and GEOBOL, La Paz, Bolivia. In addition, the following
collections were examined: in Argentina: Museo de La Plata (MLP), Museo de Ciencias Naturales
‘Bernardino Rivadavia’ of Buenos Aires (MACN), Museo Municipal de Ciencias Naturales ‘L.
Scaglia’ of Mar del Plata (MMCN), Museo de Historia y Tradicion of Loberia (MHTL), Institute
Miguel Lillo ofNational University ofTucuman (IML); in Colombia: Museo de Ciencias of Lasalle
University (MCLU) and National University of Bogota (NUB); in USA, Frick Collection of the
American Museum of Natural History of New York (AMNH); and in England, The Natural
History Museum, London (BMNH).

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

46!

Several South American specimens lack a precise stratigraphical location because they come from
very old collections where the stratigraphy is rarely indicated. Specimens analysed here come from
the following sites: Tarija (Bolivia), considered Middle Pleistocene (MacFadden et al. 1983) (7 in
Text-fig. 1); La Carolina and Salinas Oil Fields in Santa Elena Peninsula (1 and 2 in Text-fig. 1),
several localities in the Ecuadorian Andes (3 and 5 in Text-fig. 1 ) and Rio Chiche (4 in Text-fig. I ).
The Santa Elena sites are assigned to Pichilingue Formation by 14C about 26000± 100 BP (Baldock
1982), the Ecuadorian Andes localities (Chalan, Alangasi, El Colegio, Q. Colorada, Oton, Q.
Grande, etc), are placed between 40000 and 4000 years BP by Dugas (1986). The Rio Chiche site
is considered to be older than the others (Hoffstetter 1952), and all localities from Brazil (8 to 12
in Text-fig. 1) and Buenos Aires province (13 to 19 in Text-fig. 1) are considered to be latest
Pleistocene (Cunha 1971, 1981; Paula Couto 1979; Tonni et al. 1985; Bargo et al. 1986; Prado et
al. 1987; Alberdi et al. 1989; Guerin 1991 ; Carbonari etal. 1992). Finally, Cerrogordo in the Bogota
savanna (Colombia) is referred to the late Pleistocene by Porta 1960 (6 in Text-fig. 1).

The upper and lower dental rows, MCIII, CA, AS, MTI1I and 1FALIII from each locality or
group of localities are considered as a single Operational Taxonomic Unit (OTU). The
morphometric characters used in the analyses are after Eisenmann et al. 1988 (see Text-figs 2^4, and

text-fig. 2. The measurements of the occlusal view of the lower and upper cheek teeth, a, lower; 2, occlusal
length; 3, length of the preflexid; 4, length of the double knot; 5, length of the postflexid; 6, maximal breadth.
b, upper; 2, occlusal length; 3, occlusal length of the protocone; 4, occlusal breadth.

Table 1). The dental morphological characters are shown in Text-fig. 5. All teeth and bone
measurements are taken in millimetres. Since this analysis requires that no OTU has missing values,
and only a few bones were complete, it was virtually impossible to obtain a large data set. No effort
was made to separate males and females because, in horses, the differences in size and proportions
between the sexes are lost within the general intraspecific variation (Winans 1989). With respect to
this, Eisenmann (1979r?) analysed the morphometric variation of the metapodials of living horses
and concluded that the coefficient of variation shows a low value for all variables, except for
variable #9 in MC1I1, #8 and #9 in MTIII (Text-fig. 3). These few variabilities indicate low levels
of dimorphism.

In long bones we have also taken the slenderness index (1)

(1)

Slenderness Index (SI) =

Minimal breadth (3)x 100
Maximal length (1)

which serves as an indicator of the habitat where they live (more open or more wooded, harder or
softer ground, etc).

A total of 26 upper dental rows, 20 lower dental rows, 49 MCIII, 35 CA, 44 AS, 61 MTIII and
101 1 FA LI 1 1 have been used to establish the taxonomical structure (Appendix I). The computational

462

PALAEONTOLOGY, VOLUME 37

THIRD METACARPAL

THIRD METATARSAL

text-fig. 3. The measurements of the third metacarpal and third metatarsal.

work was done at the Laboratorio de Sistematica y Biologt'a Evolutiva (LASBE) of the Museo de
La Plata, using NTSYS-PC programs, version 1.60 (Rohlf 1991) and Statgraphic version 5.0
( 1991 ). Data were analysed by three methods : cluster analysis, principal components analysis (PCA)
and discriminant analysis (DA). Further details of these methods and computational procedures are
in Sneath and Sokal (1973), Rohlf and Bookstein (1990) and Reyment (1991). Character- by-
character correlation was obtained from each matrix by calculating the Pearson product-moment
correlation coefficient between each pair of characters in each set. These matrices served as input
in the PCA. The PCA was performed on each character-by-character correlation matrix and the first

FIRST PHALANX

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

463

text-fig. 4. The measurements of the calcaneum, astragalus and first phalanx of the third digit.

464

PALAEONTOLOGY, VOLUME 37

table I . The measurements of the limb bones as characteristics.

Measurements taken on third metacarpal and metatarsal

1 Maximal length

2 Medial length

3 Minimal breadth (near the middle of the bone)

4 Depth of the diaphysis at level of 3

5 Proximal articular breadth

6 Proximal articular depth

7 Maximal diameter of articular facet for the third carpal or tarsal

8 Diameter of the anterior facet for the fourth carpal or tarsal

9 Diameter of the anterior facet for the second carpal or tarsal

10 Distal maximal supra-articular breadth

1 1 Distal maximal articular breadth

12 Distal maximal depth of the keel

13 Distal minimal depth of the lateral condyle

14 Distal maximal depth of the median condyle

1 5 Angle subtended by condyle ends at centre of distal articulation

16 Diameter of the posterior facet for the fourth carpal

Measurements taken on calcaneum

1 Maximal length

2 Length of the proximal part

3 Minimal breadth

4 Proximal maximal breadth

5 Proximal maximal depth

6 Distal maximal breadth

7 Distal maximal depth

Measurements taken on astragalus

1 Maximal length

2 Maximal diameter of the medial condyle

3 Breadth of the trochlea (at the apex of each condyle)

4 Maximal breadth

5 Distal articular breadth

6 Distal articular depth

7 Maximal medial depth

Measurements taken on first phalanx

1 Maximal length

2 Anterior length

3 Minimal breadth

4 Proximal breadth

5 Proximal depth

6 Distal breadth at the tuberosities

7 Distal articular breadth

8 Distal articular depth

9 Minimal length of the trigonum phalangis

three factors were extracted. The character factor loadings were used to calculate the operational
unit factor scores, or projections, in the two factor spaces. To examine ordination efficiency, the
Euclidean distances between all pairs of operational units in factor space were calculated, and with
the resulting matrix the cophenetic correlation coefficient was used.

After groups were identified based on examination of the PCA, discriminant analysis was used
to establish a rule to differentiate these groups. DA, like PCA, is a linear function of the original
variable weighted by coefficient. DA also performs a rotation of the coordinate axes, but the aim

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

465

UPPER TOOTH

mesostyle

LOWER TOOTH

PP PP

Angular Rounded

DOUBLE KNOT

text-fig. 5. Morphological characters of a, the upper, and b, lower cheek teeth.

was to arrive at an orientation that maximized the separation between the identified groups. DA can
also be used to evaluate whether the centroids dilTer significantly or not, and often to identify
specimens not included in the original analysis which established the groups (Marcus 1990;
Reyment 1991).

In order to identify which character provided the best group discrimination, we chose the
characters that showed maximal contribution values for the PCA (Table 2). To recognize the

table 2. Characters that are most important for the Principal Component Analysis (PCA).

Principal
component
of limb
bones

Number of

character Eigen value

Principal
component
of limb
bones

Number of

character Eigen value

Third

Second

3

-0-566

metacarpal

1

0-213

First

1 1

0-972

2

0-209

4

0-961

Third

10

0-953

metatarsal

5

0-943

First

1 1

0-960

1

0-926

12

0-957

Second

16

-0-606

13

0-955

8

-0-452

5

0-947

14

0-247

14

0-946

6

0-240

Second

8

-0-814

7

0-232

9

-0-675

Astragalus

1

0-190

First

1

0-981

2

0190

2

0-978

5

— 0-164

7

0-973

First

Second

3

0-304

phalanx

5

-0-303

First

6

0-969

4

-0165

4

0-960

Calcaneum

1

0-957

First

1

0-939

Second

9

-0-400

6

0-935

8

-0-375

2

0-93 1

2

— 0-319

466

PALAEONTOLOGY, VOLUME 37

correlation character in each set of bones we undertook cluster analysis. The five character
correlation matrices (MCIII, CA, AS, MTIII and 1 FA LI 1 1) served as input in the calculation of a
phenogram by the unweighted pair group method, using arithmetical averages (UPGMA). The
cophenetic correlation coefficient (r) was computed as a measurement of distortion (Farris 1969).

RESULTS

Principal component analysis, based on upper and lower dentition data, permits us to identify two
main groups. The first group includes specimens of E. andium, the second group includes specimens
referred to E. insulatus , E. neogeus, E. santaeelenae and E. lasallei (Text-figs 6-7). E. andium is a
clear morphotype characterized by its small size adapted to mountain habitat, and can also be
distinguished from the others by skeletal elements (Text-figs 8-12). The second component permits
us to separate E. lasallei and E. santaeelenae and E. insulatus and E. neogeus by the occlusal breadth
of the lower cheek teeth (Text-figs 2, 7). It is important to note that differences between the various
species of Equus from South America are more evident among skeletal elements than cranial.
Moreover, differences between South American horses vary more in size than in shape,
consequently, the second component in the multivariate analysis plays a secondary role (see Table
2). The other group of species shows overlapping morphological patterns. For example, E. insulatus
is an intermediate form in size between E. andium and E. neogeus, mainly for MCIII and MTIII
(Text-figs 8-1 1). E. neogeus, from the Pampean region and several Brazilian localities, is the largest
form in this analysis. This species is differentiated from E. santaeelenae by MCIII and MTIII (Text-
figs 8, 11). On the other hand, E. santaeelenae has a 1FALIII similar to E. neogeus and MTIII close
to E. insulatus (Text-figs 11-12).

E. lasallei is known only from one skull which is close to E. neogeus in morphological
characteristics but bigger. The analysis of dentition shows little difference between them,
particularly in the occlusal breadth of the lower teeth. As we do not have skeletal elements of E.
lasallei, it is very difficult to invalidate this species.

We only have two specimens of E. martinei from Ecuador, one MTIII and one 1FALIII, that
coincide with E. insulatus in PCA (Text-figs 11-12).

Discriminant analysis based on all species provide a correct identification for five groups (Text-
figs 15-19), especially from the upper and lower dentitions (Text-figs 13-14). DA based on MTIII
and 1FALIII does not permit recognition as a group of the material comprising E. martinei (Text-
figs 18-19). Using DA on astragalus, we observed a good discrimination of the variables. They show
an overlap between E. insulatus and E. andium (weight < 4 per cent) and between E. santaeelenae
and E. insulatus (weight < 13 per cent), while E. santaeelenae has a good discrimination (100 per
cent; see Text-figure 16). From the multivariate analysis we deduce that the most characteristic
features for this discrimination are: the occlusal length of the dental specimens and occlusal breadth
of lower cheek teeth (2 and 6 in Text-fig. 2); the maximal breadths (5, 11) and all the distal maximal
depths (12, 13, 14) of the MTIII; the maximal length (1), minimal depth of the diaphysis (4) and
maximal breadths (5, 10, 11) of the MCIII; the maximal breadths (4, 6) plus maximal length (1) of
the 1FALIII (see Text-figs 3-4). It is interesting to note that most of these dimensions (all breadths
and some lengths) are, in general, closely related with the nature of the terrain (see Conclusions).

The cluster analysis in Text-figure 20 represents correlations between the different sets of
characters (MCIII, CA, AS, MTIII and 1FALIII). In the MCIII cluster, the maximal lengths (1, 2)
and the distal maximal breadths (10, 11) are highly correlated. From a functional morphological
point of view this is important because it has not only taxonomic but also biological significance.
The variations of dimensions 1 and 2 are directly related to the type of ground surface where these
horses live (lengthening of metapodials on harder ground surfaces and shortening of metapodials
on softer ground surfaces), and the variations of dimensions 10 and 11 are related with shift of
weight and gait of the animal (Sondaar 1968; Hussain 1975). In the CA cluster, the maximal lengths
(1,2) are highly correlated, while in the AS cluster the maximal length ( 1 ) and the maximal diameter
of the condyle (2) are also highly correlated. The variations of the dimensions 1 and 2 of the AS are

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

467

80-

4 0 I I I I

240 26.0 28.0 30.0 32 0

First component

text-fig. 6. Principal Component Analysis (PCA) of the upper cheek teeth. Abbreviations: A, Equus
(Amerhippus) andium ; I, Equus (Amerhippus) insulatus ; N, Equus ( Amerhippus ) neogeus ; S-E, Equus
(Amerhippus) santaeelenae; L, Equus ( Amerhippus ) lasallei.

-4 8 -

-6.4 -

£

o

o

-8.0 -

-9 6 -

S-E

-11 2 I I I I I

240 27.0 30 0 33,0 36.0

First component

text-fig. 7. Principal Component Analysis of the lower cheek teeth. Abbreviations as in Text-figure 6.

directly related to their function. The astragalus plays an important role, as a pulley, in the forward
and backward movement of the animal, since it receives the pressure of the body-weight and
transmits it to the digits. Furthermore, the variation of dimension 2 of the astragalus has great
implications in gait and speed of the horses. These functions are in turn controlled by dimensions
I and 2 of the calcaneum, which contain the insertion points of the strong muscles that contribute
to the same movement. For the MTIII cluster, the highest correlated characters are the lengths
(1, 2), proximal breadths (5, 7) and the distal depths (12, 13). For the 1FALIII cluster these are the
maximal lengths (1,2) and the proximal and distal maximal breadths (4, 6). The variations of these

468

PALAEONTOLOGY, VOLUME 37

; i i s-e/

•J

\ N
\

V 1

V i

\\

400 480 560 64.0 72.

First component

TEXT-FIG. 8

text-fig. 8. Principal Component Analysis of the third metacarpal. Abbreviations as in Text-figure 6.
text-fig. 9. Principal Component Analysis of the astragalus. Abbreviations as in Text-figure 6.

'A

~S-E-I

lS-E^-E

/MIS£S-E
rS:E \

III I
II 1

A AAA A A'

A AAA /

i/

56 0 64 0

First component

TEXT-FIG. 1 1

42,0 48 0 54.0 60 0 66.0 40.0 4b

First component

TEXT-FIG. 10

text-fig. 10. Principal Component Analysis of the calcaneum. Abbrevalions as in Text-figure 6.
text-fig. II. Principal Component Analysis of the third metatarsal. Abbrevations as in Text-figure 6.

M, E. (A.) martinei.

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

469

40

30 0

I I l

36 0 420 48 0

First component

54 0

36 - +L

I i i I , i I i i I i i I i i L

-6 -3 0 3 6 9

Discriminant function 1

TEXT-FIG. 12 TEXT-FIG. 13

text-fig. 12. Principal Component Analysis of the first phalanx of the third digit. Abbreviations as in

Text-figure 6. M, E. (A.) martinet.

text-fig. 13. Discriminant Analysis (DA) of the upper cheek teeth. Abbreviations as in Text-figure 6. +, mean

average.

-3 -

-3 1 5 9 13

Discriminant function 1

8 -

-4 -

N

+S-E

I , i

-32

. I .... I .... I
-22 -12 -2
Discriminant function 1

8

TEXT-FIG. 14 TEXT-FIG. 15

text-fig. 14. Discriminant Analysis of the lower cheek teeth. Abbreviations as in Text-figures 6 and 13.
text-fig. 15. Discriminant Analysis of the third metacarpal. Abbreviations as in Text-figures 6 and 13.

Discriminant function 2 Discriminant function

470

PALAEONTOLOGY, VOLUME 37

24

I1 I
I+

-0 6

A A

A A

+ A A
A

A A S-E

I

S-E

S-E

-2 6 -

l , , , I i , i I . i > I , i i L_

-3-1 1 3 5

Discriminant function 1

2 6 -

0 6 -A A

-04 -

'a \

+ A A

aa a

aa

A . A A
A A

+

1 1

J . i i L

-0.5 1.5 35

Discriminant function 1

5 5

TEXT-FIG. 16. TEXT-FIG. 17

text-fig. 16. Discriminant Analysis of the astragalus. Abbreviations as in Text-figures 6 and 13.
text-fig. 17. Discriminant Analysis of the calcaneum. Abbreviations as in Text-figures 6 and 13.

-5 3 -

I M

L I
S-E

I + SyE

J s4-e

AA

AA

A A aA

M*a

% *A \

A A A
A A
A

I

I +"
S-E

I

I

S-E

N-

-5

Discriminant function 1

text-fig. 18

-2 12

-2 8

-4 8 -

I I

yni

\ 'fji

1 I ,

l» t

I r .

A

A A

aA

A

A. aA At c

A1

S-E^-E
A mS-E

I i

-2 0 2 4

Discriminant function 1
TEXT-FIG. 19

text-fig. 18. Discriminant Analysis of the third metatarsal. Abbreviations as in Text-figures 6 and 13. M, E.

(A.) martinei.

text-fig. 19. Discriminant Analysis of the first phalanx of the third digit. Abbreviations as in Text-figures 6

and 13. M: E. (A.) martinei.

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

471

dimensions in the MTIII as well as 1FALIII have a biological significance similar to that outlined
for MCIII, as they reflect the skeletal modifications due to adaptation to the environment. The
1FALII1 represent an intermediate link in the type of movement and shift of weight of the horse.

SYSTEMATIC PALAEONTOLOGY

Recent papers on the Equidae in South America (Alberdi 1987; Alberdi and Prado 1989, 1992,
1993) recognize two main groups: equidiforms and hippidiforms. In general, both groups show
common features, possibly as a consequence of convergence due to their adaptation to similar
environments: (a) large skull in relation to body size where each group has a particular dental
morphology, with a certain interspecific variability; (b) a large degree of cranial flexion (Bennett
1980); (c) ventral separation of the occipital condyles; (d) robust body structure to different degrees
in the different forms. The forms are heavy and not well adapted to running.

Order perissodactyla Owen, 1848
Family equidae Gray, 1821

Subfamily equinae (Gray 1821) Steinmann and Doderlein, 1890
Genus equus Linnaeus, 1758
Subgenus equus (amerhippus) Hoffstetter, 1950

Synonymy (after Hoffstetter 1952, p. 233): = Equus ( partim ) auct. ; = Neohippus ( partim ) Abel 1913, 1914,
1919; Spillmann 1931, 1938; = Hippidium Spillman 1931, non Hippidium Burmeister 1875, nee Hippidion Owen
1869; = Amerldppus Hoffstetter, 1950.

Type-species. Equus ( Amerhippus ) andium Branco, 1883, ex Wagner, 1860.

Geographical distribution. South America.

Stratigraphical distribution. Equus ( Amerhippus ) remains come from different levels of the South American
Pleistocene, mostly from the Ensenadan and Lujanian Land Mammal Ages (Text-fig 21).

Diagnosis. Equus ( Amerhippus ) has a large skull with sharp and marked supraoccipital crest. It is
large in relation to the postcranial skeleton, and shows a narrow and lightly excavated preorbital
and nasal region. In general, there is a ventral separation of the occipital condyles but sometimes
they are joined. It has a peculiar vomer disposition, which reaches the palatal processes of the
maxillary anterior to the palatine. The upper cheek teeth have triangular protocones. The protocone
shows the distal part longer than the mesial one, and in some cases there are enamel wrinkles. Pre-
and postfossettes in the upper cheek teeth have developed folds (see Text-fig. 5a). The mandible is
robust and the double-knot in the lower teeth, the metaconid-metastylid, is rounded and angular
respectively. The linguaflexid is, in general, shallow and more angular in P3-P4 and more open in
M 1-M2. The ectoflexid varies from deep to shallow and never connects with the linguaflexid (Text-
fig. 5b). The appendicular skeleton shows a shortening of the distal part of the extremities, but not
as much as the Hippidion , and a more accentuated metatarsal distal flexion. In general, all species
have robust metapodials and the SI varies within the limits for robustness of this subgenus.

Discussion. Hoffstetter (1950 p. 433; 1952 p. 245) justified this subgenus, based on only one
characteristic : lack of infundibular marks in the lower incisor surface and consequent loss of surface
enamel. Nevertheless, Eisenmann (1979/6) analysed the first characteristic in living and fossil equids,
and concluded that they show a high variability. In our opinion, this is a very variable feature
because it is linked to the changes of the dental occlusal surfaces (Alberdi 1974). Consequently, its
systematic value is difficult to evaluate. Notwithstanding, we think it is correct to use the subgenus

472

PALAEONTOLOGY, VOLUME 37

0360 0 640 0.720 0.800 0.880 0 960 1 040

0.650 0 700 0 750 0 800

CALCANEUS

r = 0923

- [

0 860

0 880

0.900 0 920 0 940

ASTRAGALUS
r= 0.818

r

text-fig. 20. Cluster Analysis of each set of characters: third metacarpal, astragalus, calcaneum, third

metatarsal and first phalanx of the third digit.

Amerhippus for different South American species of Equus , because all groups show the same dental
morphology with a peculiarly large skull in relation to body structure, which in turn is characterized
by shortness and robustness of the extremities.

Hoffstetter (1952) thought the cubitus to be stronger in South American horses than in other
Equus and, in general, is also stronger in all similar Equus forms up to the present except E. andium
due to its small size, but this is not so clear.

The multivariate analysis permits us to distinguish five different groups of E. ( Amerhippus ), which
we ascribe to different species.

Equus ( Amerhippus ) andium Branco, 1883 ex Wagner 1860

I860 Equus fossilis andium Wagner, p. 336. [ nomen nudum after Mones (1986)].

1875 Equus quitensis Wolf, p. 155. [nomen nudum after Mones (1986)].

1931 Hippidium jijoni Spillmann, p. 50.

1938 Neohippus rivadeneira Spillmann, p. 386, fig. lc.

1938 Neohippus postremus Spillman, p. 389, fig 1e.

1992 Equus andium Branco 1883; nec Wagner-Branco Azzaroli p. 134, fig. 3 b.

Type material. Hoffstetter ( 1952, p. 247) stated the material described by Spillmann was lost, and proposed two
‘lectotypes’: a specimen of Neohippus rivadeneira (MEPN V-78) and one of Neohippus postremus (MEPN
V-430). We think it is best to consider the first as a neotype.

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

473

text-fig. 21. Scheme of the strati-
graphical levels of Land Mammal Ages
from South America (SALMA): Cha-
padmalalan, 'Uquian', Ensenaden and
Lujanian (modified from Tonni et al.

1992).

PALEOMAGNE-
TIC SEQUENCE

SALMA

Lujanian

Ensena

dan

Chapad •
malalan

Geographical distribution. From Oton, in the North, to Punin in the South of Ecuardorian Andes (Ecuador).
Stratigraphical distribution (Text-Fig. 21). The Cancagua Formation (Sauer 1965) from Ecuador; cited as
Upper Pleistocene after Eloffstetter (1952), although MacFadden and Azzaroli (1987) placed it between Middle
and Upper Pleistocene. Dugas (1986) described it as between 40000 and 4000 years BP and Azzaroli (1992)
referred it to Late Pleistocene.

Type stratigraphical level. ‘Puninian’ (Hoffstetter 1952, p. 248). Lujanian Land Mammal Age.

Studied material. Material studied by Hoffstetter (1952) from several localities of the Ecuadorian Andes, stored
in MEPN and IPMNHN, and the material deposited in the American Museum of Natural History (Frick
Collection).

Diagnosis. As for the subgenus Amerhippus , this species has a wide, low and laterally placed orbit.
Particularly, the skeleton of E. andium is characterized by short and robust limbs, most significantly
in the radius and metapodials, thus producing unusual proportions between these. It correspbnds
to a morphotype easily distinguishable from the others by the multivariate analysis of the dental and
limb bones (Text-figs 6-12).

Discussion. This species has been described by Branco (1883) and Hoffstetter (1952). This form
possibly presents an anatomical adaptation to environmental factors, reflected in the metapodial
shortness (Mean slenderness index : MCIII = 18,27; MTIII = 15,81). Hoffstetter ( 1 952) expressed
doubts about this relation, considering the shortness of the extremities to be differential. This
species, adapted to the Ecuadorian Andes, probably evolved from a larger form such as E. in.su/atus.
It is the smallest form of Equus from South America.

474

PALAEONTOLOGY, VOLUME 37

Equus ( Amerhippus ) insulatus C. Ameghino, in F. Ameghino 1904

1851 Equus macrognathus Weddell, p. 204.

1855 Equus neogaeus Lund; Gervais, p. 33, pi. 7, figs 2-3.

1904 Equus insulatus C. Ameghino [unpublished]; F. Ameghino, fig. 250.

1938 Neoltippus martinet Spillmann, p. 382, fig. 1b.

1952 Equus ( Amerhippus ) martinei (Spillmann, 1938); Hoffstetter, p. 301.

1992 Equus martinei Spillmann, 1938; Azzaroli, p. 137.

Holotype. MACN 1703. upper cheek teeth (M3 nor M2) figured by F. Ameghino (1904, p. 190, fig. 250). This
specimen is not lost as suggested by Mones (1986).

Geographical distribution. Tarija (Bolivia), and the Ecuadorian Andes (Rio Chiche, Ecuador).

Stratigraphical distribution. (Text-fig. 21). The Tarija localities (MacFadden et al. 1983), and the Rio Chiche
locality (Sauer 1965) are referred to the Middle Pleistocene. Hoffstetter (1952) considered E. martinei to be the
oldest Equus material from Ecuador; Clapperton and Vera (1986); followed by Azzaroli (1992), assigned this
locality to the Late Pleistocene.

Type stratigraphical level. Ensenadan Land Mammal Age.

Studied material. Includes material studied by Boule and Thevenm (1920) and MacFadden and Azzaroli (1987)
from Tarija (Bolivia) and that from Rio Chiche, Ecuador (Hoffstetter 1952).

Diagnosis. This species has a bigger skull than Equus (Amerhippus) andium , but is similar in general
morphology. Prominent cranial flexion between face and braincase. Nuchal crest extends posterior
to occipital condyles. External auditory meatus located close to glenoid fossa. The preorbital region
is also narrow but less excavated. Mandible deep and massive. Upper dental pattern characteristic
of Equus but larger in size. Protocones moderately elongated and fossettes moderately plicated.
Ectoflexids relatively shallow in the premolars and relatively deep in molars. The body size is
intermediate between E. (A.) andium and the other species studied here (see Boule and Thevenin
1920).

Discussion. Limb bones correspond to a robust horse, with slenderness index of metapodials:
MC1II = 18,16; MTIII = 16,01. E. insulatus , from Tarija, is similar in skull size to E. neogeus
from Buenos Aires province, however it differs from E. neogeus in that it has dolichocephalic skull
and a relatively high and narrow rostrum. The multivariate analysis permits a clear discrimination
of this species from the others. The scarce specimens assigned to E. (A.) martinei by Spillmann
(1938) and Hoffsteiuer (1952) are reminiscent of E. curvidens, but they have different skeletal
proportions. We think they can be ascribed to E. (A.) insulatus (Text-figs 1 1-12) by the multivariate
analysis and the morphological characters. DA also confirm this determination. The stratigraphical
distribution of E. (A.) insulatus , classically considered as Middle Pleistocene, possibly extends to the
Late Pleistocene according to some authors.

Equus (Amerhippus) neogeus Lund, 1840

1840 Equus neogeus Lund, p. 319.

1840 Equus ; Owen, p. 108, pi. 32, figs 13-14.

1845 Equus curvidens ; Owen, p. 235.

1875 Equus argentinus Burmeister, p. 55, p. 4, fig. 6.

1880 Equus rectidens Gervais and Ameghino, p. 92.

1881 Equus lundii Boas, p. 307, pis 1-2; [figs 10-20 grouped all equidiform material from the Lagoa
Santa].

1905 Equus haasei Reche, p. 225, pi. 22.

1912 Equus neogaeus [.vie] Lund 1840; Sefve p. 138.

1981 Equus (Amerhippus) vandonii Cunha, p. 5, pis 1-3.

1987 Equus (Amerhippus) curvidens Owen; MacFadden and Azzaroli p. 331.

PRADO AND ALBERD1 : SOUTH AMERICAN EQUUS

475

1992 Equus curvidens Owen; Azzaroli p. 134, fig. I b.

1992 Equus neogeus Lund; Azzaroli, p. 134 [recorded as an uncertain species].

Holotype. Specimen 866, a right metacarpal III in the Peter W. Lund Collection, Zoologisk Museum,
Copenhagen, Denmark.

Geographical distribution. Principal remains are from the Pampean region, Argentina (see Studied material) and
others from Lagoa Santa (Lund 1840), Corumba (Cunha 1981 ), Sao Raimundo Nonato, Piaui (Guerin 1991 )
and Chique-Chique and Aguas do Araxa (Paula Couto 1979), in Brazil.

Stratigraphical distribution. Upper Pleistocene of Buenos Aires province, Argentina, and Brazil.

Type stratigraphical level. Lujanian Land Mammal Age.

Studied material. This includes the material studied by Sefve (1912), such as E. neogaeus and E. curvidens from
Mercedes (Lujan), Ayacucho, Necochea, Rio Quequen Salado, Paso Otero, Arroyo Tapalque, Monte-
hermoso, among others, in Buenos Aires province, dated by 14C between 28000 and 4000 years BP (Carbonari
et al. 1992), and the material from Brazilian localities.

Diagnosis. This is one of the largest species of South American horses. The skulls are big and show
an enlarged preorbital and nasal region. The limb bones are large and robust, but more slender than
in the other South American Equus species. See the description by Sefve (1912, p. 138).

Discussion. The multivariate analysis both distinguishes this species from the rest (Text-figs 8,
1 1-12), and groups together all specimens from the Buenos Aires province sites, and the Brazilian
localities. They are slenderer (MCIII = 16, 16; MTII1 = 12, 33). Sefve (1912) thought it very
difficult to separate this species, if its dimensions are not considered, and explained that E. neogeus
is both the biggest and the most slender South American Equus. Winge (1906) synonymized E.
neogeus with E. curvidens , but we consider that priority belongs to E. neogeus.

From the nomenclatorial point of view, Lund (1840) described E. neogaeus from a metatarsal III
found at Lagoa Santa; in 1846, he described new remains from the same place. He referred two
molars to E. neogaeus , one molar to E. principalis and the rest to E. aff. caballus. Gervais (1855)
assigned part of this material to E. neogaeus and part to E. devi/lei. Owen (1869) erected Hippidion ,
which in 1870 included E. neogaeus , E. principalis and E. arcidens. Boas (1881) considered that the
metatarsal described by Lund (1840), to be a metacarpal, and therefore he created a new species,
E. lundii ; the molars referred by Lund ( 1846) as E. aff. cabal/us were considered conspecilic. Winge
(1906) considered all material from Lagoa Santa as E. curvidens. Sefve (1912) referred to E. neogaeus
as all the material from Lagoa Santa and to E. curvidens as the material from the Pampean region.
In 1971, Cunha summarized these references thoroughly and explained that only the teeth referred
by Lund (1846) to E. neogaeus and E. principalis , correspond to Hippidion , while the metacarpal and
remaining teeth correspond to Equus.

Owen (1840) mentioned and figured one horse which he later (1845) named E. curvidens. Based
on the rules of Principle of Priority (ICZN, 1985), we consider Equus neogeus (not neogaeus ) should
have priority over E. curvidens.

Equus (Amerhippus) santaeelenae (Spillmann, 1938)

1938 Neohippus santae-elenae Spillmann, p. 384, fig. Id.

Neotype. MEPN V-3037, partial skull of an adult male, Hoffstetter (1952 fig. 85a). The skull figured and
photographed by Azzarolli (1992 p. 138, text-figs 1 d and 2 is specimen V-3037, not V-68).

Paratvpes. V-l, V-3, V-5-V-6, V-IO. V-12, V-18, V-20, V-23, V-25-V-30, V-35, V-37, V-40, V-44, V-52, V-59,
V-63, V-65, V-68-V-69, V-175-V-180, V-182-V-183, V-187, V-191, V-192, V-215-V-216, V-224. V-242.
V- 1 402- V- 1404, V-l 407, V-1457, V-I460-VI462, V-3798.

Geographical distribution. La Carolina and Salinas Oil Fields localities in Santa Elena peninsula (Ecuador).

476

PALAEONTOLOGY, VOLUME 37

Stratigraphical distribution. Upper Pleistocene. Material was recovered from Pichilingue Formation, dated by
1JC as 26 000 ± 100 BP (Baldock 1982).

Type stratigraphical level. Lujanian Land Mammal Age.

Studied material. Material from the Frick collection, [AMNHJ; material studied by Spillmann (1938) and
Hofifstetter (1952) from MEPN and MNHN.

Comparative diagnosis. Mandible robust similar to the other species of E. (Amerhippus) and with
more posterior position of the canine. The molars are proportionately wider in relation to their
length. In the upper cheek teeth, the enamel is more wrinkled than in E. andium , while in the lower
cheek teeth, the wrinkling is more complex. The postcranial skeleton is larger and stronger than in
E. andium , but similar in morphology despite living in a different environment. The shortening of
the length of the radius and metapodials are also similar to those observed in E. andium. However,
E. santaeelenae , has a wider and heavier skeleton: see Spillmann (1938, p. 384) and Hofifstetter
(1952, p. 288).

Discussion. The skull fragment is similar to Equus (Amerhippus) neogeus , but smaller. The protocone
is longer, elongated at the distal part, with a certain enamel complication and the lingual groove
very marked (Hofifstetter 1952, pp. 288-301). Only the multivariate analysis of lower dentition
shows a clear discrimination for this species (Text-tig. 7). The analysis of the postcranial bones
shows that the MTIII and AS are like E. insulatus morphology (Text-figs 9 and 1 1), while 1FALIII
overlaps E. neogeus (Text-fig. 12). This indicates different proportions in relation to the other forms.
One particular feature in this species is the slenderness index that shows different patterns between
MCIII (18,18) and MTIII (15,1 1). On the other hand, the similarities between E. santaeelenae and
E. andium observed by Hofifstetter (1952) have not shown up in the multivariate analysis carried out
by us on MCIII, CA, AS MTIII, and 1 FALIII, where the studied material has been more
abundant.

We think that E. (A.) santaeelenae shares some adaptive characters in limb bones with E. neogeus ,
but differs from it in dental morphology, especially in the protocone.

Equus (Amerhippus) lasallei Daniel, 1948
1948 Equus lasallei Daniel, p. 278, fig. 66.

1960 Equus (Amerhippus) lasallei Daniel; Porta, p. 53, figs 3-8, pis 1-2.

Holotype. MCLU, unnumbered specimen, an old skull from Cerrogordo, Colombia (Daniel, 1948).
Redescribed and figured by Porta (1960, p. 53, pis 1-2).

Geographical distribution. Cerrogordo (Porta 1960) and Tibito (Correal Urrego 1981), Colombia.

Stratigraphical distribution. Late Pleistocene. Porta (1960) correlated the Cerrogordo locality with the Punian
in Ecuador (sensu Hofifstetter 1952). The Tibito site was dated by 14C as 1 1 740 ± 1 10 years BP (Correal Urrego
1981).

Type stratigraphical level. Lujanian Land Mammal Age.

Studied material. A complete skull collected by Brother Daniel (Porta 1960) and a few remains from Tibito
(Correal Urrego 1981) from MCLU and NUB (Colombia).

Diagnosis. The skull is high and long, with a longer diastema and relatively slender rostrum. The
forelobes of the occipital condyles are joint. Upper cheek teeth contain widely developed fossetes

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

477

and the enamel line is wrinkled giving a complex pattern. Lower cheek teeth are greater in breadth
than in other species. See Porta (1960, p. 253).

Discussion. Equus lasallei has the largest skull among the South American horses. The presence of
the forelobes of the occipital condyles joint is also observed in the skull described by Reche (1905)
as E. haasei. Both skulls belong to old animals. We have observed the joint to be a variable character
in some equids, and consequently we think it is possible that it may be related to the age of the
animal. Unfortunately, we know only the skull of this species, which corresponds to a male, and do
not yet have limb bones to permit more extensive analysis. The multivariate analysis, which was
clearly taken only for dentition, shows that the lower cheek teeth produce the best discrimination
(Text-fig. 7) and the breadths are the most significant character that contribute to the second
component. We think this Equus can be considered a good species based on the skull morphology.
This morphology is similar to E. neogeus but bigger in size. Future finds of postcranial remains
could permit us to form a better definition of this species.

CONCLUSIONS

The results of the morphological study of the skull agree with those obtained from the multivariate
analysis of the limbs. The conclusions can be summarized as follows:

Equus ( Amerhippus ) Hoffstetter (1950) arrived in South America from North America after the
Great American Biotic Interchange. This subgenus has been recorded in South America from the
Middle Pleistocene to the Pleistocene-Holocene boundary.

We recognize five species of Equus ( Amerhippus ) as valid: andium Branco, 1883, ex Wagner
(1860); insulatus Ameghino, 1904; neogeus Lund, 1840; santaeelenae (Spillmann, 1938), and lasallei
Daniel, 1948. We include as synonyms: Equus curvidens Owen, 1845, E. haasei Reche, 1905 and E.
martinei (Spillmann, 1938) from the latest review by Azzaroli (1992).

The geographical distributions of these species are clear, without geographical overlap, from
Colombia in the north to Buenos Aires province in the south.

The oldest record is E. (A.) insulatus , from the Middle Pleistocene of Tarija (Bolivia). The other
species come from the Upper Pleistocene of different localities in Brazil, Colombia, Ecuador and
Argentina. The most southern distribution corresponds to E. neogeus in the Pampean region.

The family Equidae is classically considered as a good biostratigraphical and palaeoecological
marker based on their wide record in the Northern Hemisphere. The South American equids reflect
the last step of this evolutionary lineage in a short period of time (Pliocene/Pleistocene).
Nevertheless, the adaptive characteristics of the Equus ( Amerhippus ) species to different habitats can
be correlated with the differences in their extremities, mainly in their distal part as observed in other
Equus species (Gromova 1949; Eisenmann 1979a, 1979c, 1984; Eisenmann and Karchoud 1982;
Eisenmann and Guerin 1984). The metapodials of Equus , in general, vary in relation to the type of
ground, both in length and width, though without any direct relation between these dimensions.
Duerst ( 1926) considered the degree of slenderness of the skeleton to be related to the environmental
conditions. For example, he considered that robustness (short and wide metapodials) was related
to a humid and cold climate in a wooded environment, while conversely, slenderness (long and
narrow metapodials) was related to a warm and dry climate and open landscape. On the other hand,
Gromova (1949) established a correlation between the widening of the third phalanges of the third
digit and the characteristics of the ground; the largest are related to soft soils and the narrowest,
to hard soils. This pattern is based on the ecology of living horses. In South America, the genus
Equus is conservative in morphology and shows important differences in size and gracility of the
distal part of the extremities, according to the environmental conditions. E. (A.) neogeus and E. (A.)
santaeelenae , from the plains regions, are the largest in size; the former is more slender and comes
from the open plains; the latter, stronger, comes from a restricted area with more sandy ground.
E. (A.) insulatus and E. (A.) andium are intermediate forms with bigger and stronger second and
fourth lateral metapodials, and show comparatively shorter metapodials than the plains forms. This

478

PALAEONTOLOGY, VOLUME 37

is more marked in the last species E. andium. We consider this character as adaptively related to a
mountain habitat with hard ground and steep relief. Moreover, we think it may be considered as
an adaptive convergence with the small Hippidion species also from South America. In general, we
observed a similar structure in both Hippidion and Equus genera from South America that may
indicate an adaptive convergence to the environment. In work in progress we are attempting to
determine the relation between body size and weight of Equidae from different continents.

We conserve the validity of name E. (A.) lasallei for the Cerrogordo (Colombia) specimens while
awaiting further evidence of the postcranial skeleton.

Acknowledgements. The authors thank the curators of the collections in different museums and institutes for
kindly allowing us to study material. We wish to express our thanks to Drs E. Tonni and E. Cerdeno for their
critical revision of the manuscript and valuable discussions. We also thank Dr A. R. Milner, Dr M. Fortelius
and one anonymous referee for their criticism, comments and suggestions. The drawings were prepared by
J. Arroyo and some of the tables typed by M.T. Montero (MNCN). The present work was aided by a
Research Grant CE num. Cl 1 *-CT90-0862, the Spanish Grants DGICYT PB-88-0008 and PB-91-0082, and
the Bilateral Convention CSIC-CONICET.

REFERENCES

alberdi, m. t. 1974. El genero Hipparion es Espana. Nuevas formas de Castilla y Andalucta, revision e historia
evolutiva. Trabajos sobre Nedgeno-Cuaternario, CSIC, 1, 1-146.

1987. La Familia Equidae, Gray 1821 (Perissodactyla, Mammalia) en el Pleistoceno de Sudamerica.
IV Congreso Latinoamericano de Paleontologia, Bolivia 1, 484-499.

— menegaz, a. n., prado, J. l. and tonni, e. p. 1989. La fauna local de Quequen Salado-Indio Rico
(Pleistoceno tardi'o) de la provincia de Buenos Aires, Argentina. Aspectos Paleoambientales y
Biostratigraficos. Ameghiniana , 25, 225-236.

— and prado, j. L. 1989. Multivariant Analysis of remains of the appendicular skeleton of South American
Equids. Abstract of Papers and Posters , Fifth Internationa! Theriological Congress , Rome , August 1989. 1,
261-262.

1992. El registro de Hippidion Owen, 1869 y Equus (Amerhippus) Hoffstetter, 1950 (Mammalia,
Perissodactyla) en America del Sur. Ameghiniana , 29, 265-284.

1993. Review of the genus Hippidion Owen, 1869 (Mammalia: Perissodactyla) from the Pleistocene
of South America. Zoological Journal of the Linnean Society , 108. 1-22.

ameghino, F. 1904. Recherches de morphologie phylogenetique sur les molaires superieures des ongules. Anales -
Museo National, Buenos Aires, 3, 541.

azzaroli, a. 1992. The present status of knowledge on the Ecuadorian species of the genus Equus. Bollettino
della Societd Paleontologica Italiana, 31, 133-139.

baldock, j. w. 1982. Geologia de Ecuador. Boletin de la Explicacion de Mapa Geologico de la Reptiblica del
Ecuador Escala 1:1000000. Division de Investigacion Geologica-Minera, 66 pp.

BARGO, M. S., MENEGAZ, A. N„ PRADO, J. L., SALEMME, M. C., TAMBUSSI, C. and TONNI, E. P. 1986. MamiferOS y
bioestratigrafia. Una nueva fauna local de la Unidad Mamifera Lujanense (Pleistoceno tardio) de la
provincia de Buenos Aires. Ameghiniana 23, 229-232.

bennett, d. k. 1980. Stripes do not a zebra make. Part I: A cladistic analysis of Equus. Systematic Zoology,
29, 272-287.

boas, J. e. v. 1881. Om en fossil Zebra-Form fra Brasilienes Campos. Med. et Tillaeg om to Alter af Slaegten
Hippidion. Videnskabernes Selskabs Skrifier , 6. Raekke , naturvidenskabelig og mathematisk Af deling, 1,
305-330.

boule, m. and thevenin, a. 1920. Mammiferes fossiles de Tarija. Paris Imprimerie Nationale, 256 pp.

branco, w. 1883. Ueber eine fossile Saugethier-Fauna von Punin bei Riobamba in Ecuador. IE Beschreibung
der Fauna. Palaeontologische Abhandlungen, 1, 57-204.

burmeister, G. 1875. Los coba/los fosi/es de La Pampa Argentina. La Tribuna, Buenos Aires, 88 pp.

carbonari, j. e., huarte, r. a. and figini, a. j. 1992. Miembro Guerrero, Formation Lujan (Pleistoceno,
provincia de Buenos Aires) edades Cl 4, Jornadas Geoldgicas Bonaerenses, 245-247.

clapperton, c. m. and vera, r. 1986. The Quaternary glacial sequence in Ecuador: a reinterpretation of the
work of Walter Sauer. Journal of Quaternary Research, 1, 45-56.

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

479

correal urrego, G. 1981. Evidencias culturales y megafauna pleistocenica en Colombia. Fundacion de
Investigaciones Arqueologicas Nacionales. Banco de la Repubhca, Bogota, Colombia, 12, 148 pp.
cunha, f. l. de s. 1971 . Sobre os holotipos de Equus (. Amerhippus ) neogaeus Lund, 1840 e Equus ( Amerhippus )
curvidens Owen, 1844. Anais de Academia Brasileira de Ciencias, 43 (supplemento), 619-627.

— 1981. Equus (Amerhippus) vandonii n. sp. Um novo cavalo fossil de Corumba, Mato Grosso do Sul, Brasil.
Boletim do Museo National, ( Geologia ), 40, 1-19.

daniel, h. 1948. Nociones de Geologia y Prehistoria de Colombia. Medellin, 360 pp.

duerst, j. u. 1926. Vergleichende Untersuchungen am Skelett bei Saugern. Arbeitsmethodologie Abderhalden ,
7, 200.

dugas, F. 1986. Historico Cientifico de la Quebrada Chalan-Area de Panin, Provincia de Chimborazo. Proyecto
EPN-IPGF1-ORSTON. Facultad de geologia, Minas y Petroleos. Escuela Politecnica Nacional, Quito,
15 pp.

eisenmann, v. 1979a. Les metapodes d 'Equus sensu lato (Mammalia, Perissodactyla). Geobios, 12, 863-886.

— \919b. Etude des cornets des dents incisives inferieures des Equus (Mammalia, Perissodactyla) actuels et
fossiles. Palaeontographia Italica, 71, 55-75.

— 1979c. Les chevaux (Equus sensu lato) fossiles et actuels: etude craniologique et odontologique. Ph.D.
thesis, Universite Pierre et Marie Curie, Paris.

— 1984. Sur quelques caracteres adaptatifs du squelette d 'Equus (Mammalia, Perissodactyla) et leurs
implications paleoecologiques. Bulletin Museum national d Histoire naturelle, Paris, Section C, 6, 185-195.

— alberdi, M. T., de giuli, c. and staesche, u. 1988. Studying fossil horses, 1, Leiden, E. J. Brill, 72 pp.

— and guerin, c. 1984. Morphologie fonctionnelle et environnement chez les Perissodactyles. Geobios,
Memoire special, 8, 69-7 4.

— and karchoud, a. 1982. Analyses multidimensionnelles des metapodes d 'Equus (Mammalia, Peris-
sodactyla). Bulletin Museum national d Histoire naturelle , Paris, Section C, 4, 75-103.

farris, J. s. 1969. On the Cophenetic Correlation Coefficient. Systematic Zoology, 18, 279.
gervais, p. 1855. Recherches sur les Mammiferes fossiles de I'Amerique meridionale. P. Bertrand, Libraire-
Editeur, Paris, 63 pp.

— and ameghino, f. 1880. Les Mammiferes fossiles de I'Amerique du Sud. Paris et Buenos Aires, 225 pp.
gromova, v. 1949. Histoire des chevaux (genre Equus) de l’Ancien Monde. Trudy Paleontologicheskogo

Instituta Akademiya Nauk SSSR , 7, 1-373. [French translation : Centre Etudes Documents Paleontologie 13,
Paris 1965],

guerin, c. 1991. La faune de vertebres du Pleistocene superieur de l’aire archeologique de Sao Raimundo
Nonato (Piaui, Bresil). Comptes Rendus de I'Academie des Sciences , Paris serie II, 312, 567-572.
hoffstetter, r. 1950. Algunas observaciones sobre los caballos fosiles de la America del Sur. Amerhippus gen.
nov. Boletin Informations Cientificas Nacionales, 3, 426-454.

— 1952. Les mammiferes Pleistocenes de la Republique de l’equateur. Memoires de la Societe Geologique de
France, 31. 1-391.

hussain, s. t. 1975. Evolutionary and functional anatomy of the pelvic limb in fossil and Recent Equidae
(Perissodactyla, Mammalia). Anatomy, Histology, Embryology, 4, 179-222.
iczn, 1985. International code of zoological nomenclature, 3rd edition. University of California Press, 338 pp.
lund, p. w. 1840. Nouvelles recherches sur la faune fossile du Bresil. Annales des Sciences Naturelles, 13,
310-319.

— 1846. Meddelelse af det Udbytte de i 1844 undersogte Knoglehuler Have avgivet til hundskaben om
Brasiliens Dyreverden for sidste Jordomvaeltning. Det kongelige Danske Videnskabernes Selskabs
naturvidenskabelige og mathematisk Afhandlinger , 12, 57-94.

macfadden, b. j. and azzaroli, a. 1987. Cranium of Equus insulatus (Mammalia, Equidae) from the Middle
Pleistocene ofTarija, Bolivia. Journal of Vertebrate Paleontology, 7, 325-334.

— siles, o., zeitler, p., Johnson, n. m. and Campbell, k. e. 1983. Magnetic polarity stratigraphy of the
Middle Pleistocene (Ensenadan) Tarija Formation of southern Bolivia. Quaternary Research, 19, 172-187.

marcus, l.f. 1990. Traditional morphometries. In rohlf, f. j. and bookstein, f. l. (eds). Proceedings of the
Michigan Morphometries Workshop, University of Michigan, Museum of Zoology, Special Publication, 2,
78 122.

mones, a. 1986. Palaeovertebrata Sudamericana. Catalogo sistematico de los Vertebrados Fosiles de America
del Sur. Parte I. Lista preliminar y Bibliografia. Courier Forschungsinstitut Senckenberg, 82, 1-625.

OWEN, R. 1840. The zoology of the voyage of H.M.S. Beagle under the command of Captain Fitzroy, R.N., during
the years 1832 to 1836. Part 1 Fossil Mammalia. Ed. y superv. C. Darwin. 1 1 1 pp.

480

PALAEONTOLOGY, VOLUME 37

OWEN, R. 1845. Descriptive and illustrated catalogue of the fossil organic remains of Mammalia and Aves
contained in the Museum of the Royal College of Surgeons of England , R. and J. E. Taylor, London, 391 pp.

1869. On fossil teeth of Equines from Central and South America, referable to Equus conversidens, Equus
tail , and Equus arcidens. Proceedings of the Royal Society of London, 17, 267-268.

- 1870. On fossil remains of Equines from Central and South America referable to Equus conversidens , Ow.,
Equus tau , Ow., Equus arcidens , Ow. Philosophical Transactions of the Royal Society of London , 159.
559-573.

paula couto, c. 1979. Paleomastozoologia. Academia Brasileira de Ciencias, Rio de Janeiro, 590 pp.
porta, J. de 1960. Los equidos fosiles de la sabana de Bogota. Boletin de Geologia. Universidad Industrial de
Santander , Colombia, 4, 51-78.

prado, j. L., menegaz, a. n., tonni, E. p. and salemme, m. c. 1987. Los mamiferos de la fauna local Paso Otero
(Pleistoceno tardio) provincia de Buenos Aires. Aspectos paleoambientales y biogeograficos. Ameghiniana ,
24, 217-233.

reche, o. 1905. Ueber eine neue Equidenart aus der Pampasformation. Beitrage zur Paldontologie und Geologie
Osterreich-Ungarn und des Orients, 18, 225-241.
reyment, R. a. 1991. Multidimensional paleobiology. Pergamon Press, 416 pp.

rohlf, f. j. 1991. NTSYS-pc numerical taxonomy and multivariate analysis system. Version 1.60. Exeter
Publication, New York.

— and bookstein, F. L. 1990. Proceedings of the Michigan morphometries workshop. The University of
Michigan , Museum of Zoology, Special Publication, 2, I 380.

sauer, w. 1965. Geologia del Ecuador, del Ministerio de Educacion, Quito-Ecuador, 383 pp.
sefve, i. 1912. Die Fossilen Pferde Siidamerikas. Kungliga Svenska Vetenskaps-Akademiens Handlingar, 48,
1-185.

sneath, p. H. a. and SOK at, r. r. 1973. Numerical taxonomy. The principles and practice of numerical
classifications. Freeman, San Francisco, 573 pp.

sondaar, p. y. 1968. The osteology of the manus of fossil and Recent Equidae with special reference to
phylogeny and function. Verhandelingen der Koninklijke Nederlandse Akademie van Wetenschappen, 25,
1-76.

Spillman, F. 1931. Die Saugetiere Ecuadors im Wandel der Zeit, 1, 1-112.

1938. Die fossilen Pferde Ekuadors der Gattung Neohippus. Palaeobiologica, 6, 372-393.
statgraphic 1991. Version 5.0. Statistical graphic corporation.

tonni, e. p., prado, J. l., menegaz, a. N. and salemme, m. c. 1985. La unidad marm'fero (Fauna) Lujanense.

Proyeccion de la estratigrafia mamaliana al Cuaternario de la region Pampeana. Ameghiniana, 22, 255-261.
wagner. A. 1860. Ueber fossile Saugetierknochen am Chimborasso. Sitzungsberichte der koniglich bayerischen
Akademie der Wissenschaften zu Miinchen, 3, 330-338.
webb, s. d. 1976. Mammalian faunal dynamics of the great American Interchange. Paleobiology, 2, 220-234.
weddell, h. a. 1851. Voyage dans le Sud de la Bolivie. In f. de castelnau (ed.). Expedition dans les parties
centrales de f Amerique du Sud, 1, 1-432.

winans, m c. 1989. 14. A quantitative study of North American fossil species of the genus Equus. In
d. r. prothero and r. m. schoch (eds). The evolution of Perissodactyls. Oxford monographs on geology and
geophysics, 15, 262-291.

winge, h. 1906. Jordfundne og nulevende Hovdyr (Ungulata) fra Lagoa Santa, Minas Geraes, Brasilien.
Sanding af Afhandlinger e Museo Lundii, 3, 1-239.

wolf, t. 1875. Geognostische Mittheilungen aus Ecuador. 4: Kritische Zusammenstellung der in Ecuador
stattgefundenen Vulkan-Ausbrtiche und Erdleben seit der Zeit der Conquista. Neues Jahrbuch fur
Mineralogie, Geologie und Paldontologie, 1875, 152-170.

JOSE L. PRADO

Museo de La Plata
Paseo del Bosque s/n
1900-La Plata, Argentina

MARIA T. ALBERDI

Typescript received 16 November 1992
Revised typescript received 21 May 1993

Museo Nacional Ciencias Naturales
CSIC

Jose Gutierrez Abascal 2
28006-Madrid, Spain

PRADO AND ALBERDI: SOUTH AMERICAN EQUUS

481

APPENDIX I

Localities of specimens included in this study grouped by species
Equus ( Amerhippus ) andium.

Upper dentition: MEPN : V.2152, V.76, V.2161, V.2135, V.2163, V.374, V.381, V.403; IPMNHN : V.2159;
AMNH : #19, #6. Lower dentition-. MEPN: V.2162, V.2138, V.2164, V.2132, V.2145, V.28E V.2186, V.290,
V.242; IPMNHN: V.2181. MCIII: MEPN: V.2280, V.24I4, V.2408, V.24I2, V.4014, V.470, V.2415, V.2411,
V.466, V.2416, V.1499, V.2301, V.1505, V.2302, V.469, V.472, V.467, V.2410, V.464, V.465, V.4084, V.1501 ;
IPMNHN: 6 unnumbered specimens, Hoffstetter Col.; AMNH: EcuA.4l, EcuA42, 5 unnumbered specimens.
CA : MEPN: V. 2388, V.2229, V.2220, V. 2281, V. 2222, V.223 1 , V.2218, V.2219, V.2386, V.2216, V.517, V.4I II,
V.2221, V.41 12, V.41 14, V.2232, V.516; IPMNHN : AA4/7, 1 unnumbered specimen; AMNH : 10 unnumbered
specimens. AS: MEPN: V.2239, V2251, V2223, V.2249, V.2236, V.51 1, V.2390, V.2286, V.510, V.507, V.508,
V.2248, V.2242; IPMNHN: 1 unnumbered specimen, Hoffstetter Coi.; AMNH: 17 unnumbered specimens.
MTIII : MEPN: V.458, V.2417, V.2328, V.1492A, V.1489, V.1490, V.1495, V.1494, V.2222, V.I493, V.1488,
V.4073, V.4076, V.456; IPMNHN: 6 unnumbered specimens, Hoffstetter Col.; AMNH: 14 unnumbered
specimens. 1FALIII: MEPN: V.2352, V.474, V.2325, V.2328, V.2346, V.2343, V.2329, V.4086, V.2339, V.2336,
V.2397, V.2338, V.2401, V.501, V.483, V.2334, V.2228, V.2327, V.2332, V.2341, V.2349, V.2335, V.2331,
V.502, V.2330, V.477, V.481, V.2342, V.2396 ; IPMNHN : 2 unnumbered specimens, Hoffstetter Col. ; AMNH :
29 unnumbered specimens.

Equus ( Amerhippus ) insulatus

Upper dentition: IPMNHN: TAR.996, TAR. 997, TAR. 783, TAR. 1 142; MNHN : 002160, 001432, 000922, 1
unnumbered specimen. Lower dentition: MNHN: 000679, 001576, 001577, 001415. MCIII IPMNHN:
TAR. 1175, TAR. 1179, TAR. 1176, TAR. 1 177, TAR. 1178. CA: MNHN: 001296, 001293, 001298, 001297;
GEOBOL: UF. 91956. AS: MNHN: 001302, 001303,001301,001306, 001305, 001291,001290,001304. MTIII:
MEPN: V.543 (E. martinei ); IPMNHN: TAR. 1192, TAR. 1 188, TAR 1 190, TAR. 1 182, TAR. 1 193; MNHN:
001020, 001029, 001031, 001028, 001032, 001018, 001019, 001010. 1FALIII: MEPN: V.543 (E. martinei );
MNHN: 001256, 001254, 001255, 001263, 001257, 001258, 001252, 001265, 001266, 001268, 001260, 001277,
001270, 001275, 001271, 001269, 001259, 001262, 001273, 001274, 003132, 001038, 001013, 003129.

Equus ( Amerhippus ) neogeus

Upper dentition: MLP: 6.1, 6.7; MACN : 1288, 11721, 11 15. Lower dentition: MACN : 9753, 2835, 1603,
5401 (1835). MCIII: MLP: 85-11- 10.3a, 63-VI-10.17, 6.305, 6.402 (6.412), 6.106, 6.306, 6.397, 28-III-16.8.
MTIII: MLP: 68-IX-3.3, 85-11- 1 0.4, 63-VI-10.17, 6.42, 6.56, 6.10; MACN: I unnumbered specimen. 1FALIII:
MLP: 85-11-10. 3b, 85-11-13. 3b, 86-III-25. 1 7, 55-VIII-12.i l; 6.42, 6.360, 6.219, 3 unnumbered specimens (PA-
OT).

Equus (Amerhippus) santaeelenae

Upper dentition: MEPN: V.3037. Lower dentition: AMNH: 39409. MCIII: MEPN: V.37. CA: IPMNHN:
LAR.55 (V.20); AMNH: I unnumbered specimen. AS: MEPN: V.3786, V.I7, V.16; IPMNHN: LAR.56
(V.18); AMNH: 1 unnumbered specimen. MTIII: MEPN: V.35, V.31 ; IPMNHN: LAR.48 (V.33); AMNH:
3 unnumbered specimens. IFALIII: MEPN: V.40, V.44, V.47, V.41, V.45, V.43; IPMNHN: LAR.51 (V.42),
LAR.65 (V.40).

Equus (Amerhippus) lasallei

One cranial specimen with upper and lower dentition (MCLU, unnumbered).

NOTES FOR AUTHORS

The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts should conform
in style to those already published in this journal, and should be sent to Dr R. M. Owens, Department of Geology, National
Museum of Wales, Cardiff CF1 3NP, UK, who will supply detailed instructions for authors on request (these are published
in Palaeontology 1990, 33, 993-1000).

Special Papers in Palaeontology is a series of substantial separate works conforming to the style of Palaeontology.

SPECIAL PAPERS IN PALAEONTOLOGY

In addition to publishing Palaeontology the Association also publishes Special Papers in Palaeontology. Members may
subscribe to this by writing to the Membership Treasurer: the subscription rate for 1994 is £55 00 (U.S. $120) for Institutional
Members, and £20 00 (U.S. $36) for Ordinary and Student Members. A single copy of each Special Paper is available on a
non-subscription basis to Ordinary and Student Members only, for their personal use, at a discount of 25 per cent below the
listed prices: contact the Marketing Manager. Non-members may obtain Nos 35-46 (at cover price) from Blackwell Publishers,
Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, UK, and older issues from the
Marketing Manager. For all orders of Special Papers through the Marketing Manager, please add £1-50 (U.S. $3) per item
for postage and packing.

PALAEONTOLOGICAL ASSOCIATION PUBLICATIONS

Special Papers in Palaeontology

For full catalogue and price list, send a self-addressed, stamped A4 envelope to the Marketing Manager. Numbers 2-39 are
still in print and are available together with those listed below:

39. (for 1988): Late Cenomanian and Turonian ammonite faunas from north-east and central Texas, by w. J. Kennedy,
131 pp., 39 text-figs, 24 plates. Price £30 (U.S. $60).

40. (for 1988): The use and conservation of palaeontological sites. Edited by p. r. crowther and w. A. Wimbledon. 200 pp.,
31 text-figs. Price £30 (U.S. $60).

41. (for 1989): Late Jurassic-early Cretaceous cephalopods of eastern Alexander Island, Antarctica, by p. j. howlett.
72 pp., 9 text-figs, 10 plates. Price £20 (U.S. $40).

42. (for 1989): The Palaeocene flora of the Isle of Mull, by m. c. boulter and z. kvacek. 149 pp., 23 text-figs, 23 plates.
Price £40 (U.S. $80).

43. (for 1990) : Benthic palaeoecology of the Late Jurassic Kimmeridge Clay of England, by p. r. wignall, 74 pp., 50 text-figs.
Price £30 (U.S. $60).

44. (for 1990): The Murchison Symposium. Edited by M. G. bassett, p. d. lane and d. Edwards. 397 pp., 137 text-figs,
3 plates. Price £60 (U.S. $120)

45. (for 1991): Coniacian ammonite faunas from the United States Western Interior, by w. J. Kennedy and w. a. cobban.
96 pp., 31 text-figs , 17 plates. Price £35 (U.S. $70).

46. (for 1991): Arundian (Lower Carboniferous) conodonts from South Wales, by J. J. stone. 63 pp., 14 text-figs, 5 plates.
Price £30 (U.S. $60)

47. (for 1992): The skeletal morphology, systematics and evolutionary history of holothurians, by p. M. Gilliland. 147 pp.,
19 text-figs, 11 plates.

48. (for 1992): Contributions to acritarch and chitinozoan research. Edited by s. G. molyneux and k. j. dorning. 139 pp.,
28 text-figs, 22 plates.

49. (for 1993): Studies in palaeobotany and palynology in honour of Professor W. G. Chaloner, F.R.S. Edited by
m. e. collinson and A. c. scott. 187 pp., 38 text-figs, 27 plates.

Field Guides to Fossils and Other Publications

These are available only from the Marketing Manager. Please add £100 (U.S. $2) per book for postage and packing plus
£1-50 (U.S. $3) for airmail. Payments should be in Sterling or in U.S. dollars, with all exchange charges prepaid. Cheques
should be made payable to the Palaeontological Association.

1. (1983): Fossil Plants of the London Clay, by m. e. collinson. 121 pp., 242 text-figs. Price £7-95 (U.S. $16) (Members £6

or U.S. $12).

2. (1987): Fossils of the Chalk, compiled by e. owen; edited by a. b. smith. 306 pp., 59 plates. Price £11-50 (U.S. $23)

(Members £9-90 or U.S. $20).

3. (1988): Zechstein Reef fossils and their palaeoecology, by n. hollingworth and t. pettigrew. iv + 75 pp. Price £4-95

(U.S. $10) (Members £3-75 or U.S. $7-50).

4. (1991): Fossils of the Oxford Clay, edited by d. m. martill and J. d. Hudson. 286 pp., 44 plates. Price £15 (U.S. $30)

(Members £12 or U.S. $24).

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $40).

1985. Atlas of Invertebrate Macrofossils. Edited by J. w. Murray. Published by Longman in collaboration with the
Palaeontological Association, xiii +241 pp. Price £13-95. Available in the USA from Halsted Press at U.S. $24-95.

© The Palaeontological Association, 1994

Palaeontology

VOLUME 37 - PART 2

CONTENTS

Architectural constraints on the morphogenesis of prismatic
structure in Bivalvia

TAKAO UBUKATA 241

Dinoflagellate cysts from the Glacial/Postglacial transition in the
northeast Atlantic Ocean

REX HARLAND 263

Ichnofabric from the Upper Jurassic lithographic limestone of
Cerin, southeast France

C. GAILLARD, P. BERNIER, J. C. GALL, Y. GRUET, G. BARALE,

J. P. BOURSEAU, E. BUFFETAUT and S. WENZ 285

The ichnogenus Beaconites and its distinction from Ancorichnus
and Taenidium

D. G. KEIGHLEY and R. K. PICKERILL 305

A large owl from the Palaeocene of France

CECILE MOURER-CHAUVIRE 339

Re-interpretation of terebratulide phylogeny based on
immunological data

K. ENDO, G. B. CURRY, R. QUINN, M. J. COLLINS, G. MUYZER

and p. westbroek 349

Middle to late Telychian (Silurian : Llandovery) graptolite
assemblages of central Wales

JAN Z ALASIEWICZ 375

The postcranial skeleton of the earliest dicynodont synapsid

Eodicynodon from the Upper Permian of South Africa

B. S. RUBIDGE, G. M. KING and P. J. HANCOX 397

A computer model for skeletal growth of stromatoporoids

ANDREW R. H. SWAN and STEPHEN KERSHAW 409

Morphology of encrusting and free living acervulinid
Foraminifera: Acervulina, Gypsina and Solenomeris

CHRISTINE PERRIN 425

A quantitative review of the horse Equus from South America

JOSE L. PRADO and MARIA T. ALBERDI 459

Printed in Great Britain at the University Press, Cambridge

ISSN 0031-0239

-*

/ s W# §

Nr k’A^trA'v^ r%

C v*

Q \w^ _ x-^>^ ^ ..»'• — x^>^ O X^osv^ “ o

Z _J 2 -I Z _j 2

VIN0SH1IINS S3 ! H VM 9 11 LIBRARIES SMITHSONIAN INSTITUTION NOUfUlISNI NVINOSHilWS S3

Z Z C“ 5^

to

z

rn x^jh-^ ^ m " xgmisgx m

MITHSONIAN“lNSTITUTION NOUfUlISNI NV1N0SH1IIAIS S3 I a VH 8 il“u B R AR I ES^SMITHSONIAN-^
z to z to z v . to z

1 1||^ 1 ^Olf I l€.J§ I Mm I *"*' « 1*3® 8 *

Vu^S/ 2 x > 2 No^dc/ -jj, 2 N >

to * z ■ . „ „ . 1/5 2 w V Z co

iviNOSHiiiMs S3 lava an libraries Smithsonian institution NouniiiSNi nvinoshiiws s;

CO '5 to ” to “ . CO

_ m «*mr ~ 0

SMITHSONIAN”" INSTITUTION N0linillSNI~NVIN0SHlllAJs""S3 lavaail^LIBRARI ES^ SMITHSONIAN^ IN
c~ z r- , z r- Z <”

o xCswoX ~ & O xSJTiroX “ ^=vcT>Sx O

pn x^<os££x " vjrAsq^ m ' 2 XjvAst^x sn x2yos*£x ^ m

JVINOSHlIIAIS^Sa I ava a ll“u B R AR I ES SMITHSONIAN ""INSTITUTION^ NOlinillSNI ~ NVINOSHlItMS S

to z „ to z » co z to

< w s XfSijytoX < U S s

^ v% . r3 ^ v% . q 5 A

x f/M'W; O

to /pw to

o SV' x

■z,"/ /S' t

^ I

o ,
z /

> S ■ XT’ > s 'NSS* >'• 2 >’

Z to , Z CO ■•* Z to Z

SMITHSONIAN INSTITUTION N0UfUllSNI_NVIN0SHllWS S3iavaail LIBRARIES SMITHSONIAN IN

_ CO ^ t. co - CO ^

§X m ^ tO V ^ Ji /^EJqX u W

O 2 O — o

_ _ z ' J 2 _j Z

iviNOSHiiiAis S3 1 ava an libraries Smithsonian institution NouniiiSNi nvinqshjjws s:

^ 1 5 1 , | 5 2

I) > (ig 3|) = > (If §| 2 ^ ^ 2 ^

M % ^§ik m w N<^^/ " ^ ^ «

^ /rt \

to

CD
73
>

73
—

m XL'* ~ m Xffi/ xcoiac^y rn

_ co _ co £ co

SMITHSONIAN INSTITUTION NOUfUlISNI NVIN0SH1IWS SBIHVaail LIBRARIES SMITHSONIAN IN

z .v co z to z w z

< S .<

\ivinoshiiws S3 1 ava an libraries Smithsonian institution NouniiiSNi nvinoshiiws s:

co = to zz to ” to

UJ

a:

<

CsL

co

.'P^ALW'A/ ij

O X^M DC ^ O x^ _ XIUCLS^ Q

SMITHSOmAN^INSTITUTION^NOlinillSNI^NVINOSHlIWS^Sa I ava 8 H^LI B RAR I ES^ SMITHSON IAN IN
r- z r= , z r° Z

o

2=> T-J) ^

III LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSH1IWS S3IHVU0n LIBRARI
t- z r~ _ ^ x r~ —

fn pcy ^ m <p x-w»iixy^/ rn ^ ^ x^jus^x m

w ~ CO — if) \ _ to

ION NOIlfUllSNI NVINOSHilWS S3IMVH9I1 LIBRARIES SMITHSONIAN INSTITUTION NOlinili!

CO X . ^ 5 .X- S2 _ 2 ,V- to

< E , v - x<mS'\ - x^SSTx -

,„/ 2 H 0\ rr .WW''

-r •./££ O /? P' X . 'v^5t O X: 2

7H •■■sffidw n il to irWAfiy/ to to

i h xi^i/ 5 t 5 \mmy t

> I 5 | i

Sn^LIBRARIES^SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHilWS^Si I B VB 8 11~ LI B RAR I
— CO ^ to X to

Q _ ><OULS2^' q

X _j X -j X *t _l X

ION NOlinillSNI NVIN0SH1IINS SBIBWaail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllS
X r~ » H r- x ^ x

O ° XZwJvZk ~ XooTS?x O m /5gTo\ O

03

>\ ” \^V, xx »- /^fr*TA ~ /KrffT F*ft\ t- ” /^-^A h-

x\ ^ '-Hu\ 33 r> •fi'A 20

li) t (t§. Ji) s vm t (ifLJj) 5

m '■ w ™ x.'>vasvix m xv\0$nj^/ ^ rn x^os>^

co ' — co _ co tz

311 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI MVIN0SH1IWS SSIBVaail LIBRAR

X , to X V CO X to X

< S x5vnro>\ < A S < XXx 2 <

£ 8

x /ip -a|\ o x ° ls£?’'rj2k'A *

s IL.JI s vS| s s If, Ji) s

CO "" X CO *■ X CO X CO

ION NOlinillSNI _NVIN0SH1IINS S3 I d VB a H_LI B RAR I ES SMITHSONSANJNSTITUTION NOlinili?
>A - 1 ^ , - /tfgw&s w w u

311 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3IBVH8I1 LIBRARI
X f“ __ X r* ^ r*

m ^ m ' ^ m

ION NOlinillSNI NVIN0SH1IWS S3 I H VB a ll~U B R AR I ES^SMITHSONIAN-INSTITUTION ^ NOlinili

CO X to X .... co X V CO

2 ,< 2 , < A,. 2 < .^S. . 2

•-/^ .

to ^
o ?S

Z

k > 2 2 XA > 2

Oil LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHlIWS^SB I H VB a II “ LI B R AR I

” CD ~ C/) ~ =

in

V —

O UV^X _ Q TT^1 >s^lllSJ^Xr Q

ION NOlinillSNI “NVINOSHilWS Saiyvaan^UBRARI ES^SMIIHSOmAN^INSTITUTION “'N0llnlll,